Sleep
May 1, 2026
Holland Doise, A. Gonzalez Quesada, Garrett Baber et al.
Nightmares are linked to more severe posttraumatic stress disorder (PTSD). This study tested whether the emotional content of dreams, measured by self-report, human coding, and a large language model (LLM), could predict PTSD symptoms after a traumatic injury. 87 patients from a trauma center provided dream reports and completed surveys during hospitalization and at one and two months afterward. Dream narratives were scored by trained raters and by an LLM for negative affect and arousal. LLM-derived negative affect and arousal from dreams reported during hospitalization predicted higher PTSD symptoms one month later, even after adjusting for other risk factors. Self-reported dream affect did not predict later symptoms.
Sleep
April 20, 2024
Garrett Baber, Daiil Jun, Jeffrey M. Girard et al.
Fearful dreams are associated with worse mood the following day, not better. Analyzing 4986 daily reports from 553 participants, the study found that on days when a person's dream contained more fear, their next-day negative affect was 14% higher—more than twice the effect of poor sleep efficiency (6%). People who generally had more fearful dreams also reported 12% higher average negative affect. These results contradict the theory that dreaming about fear helps regulate mood through a process similar to desensitization therapy. Instead, fearful dreams appear to be a meaningful day-to-day predictor of lower mood, independent of sleep quality and personality traits like neuroticism.
SLEEPJ
May 8, 2026
Andrew Meierotto, Garrett Baber, Nancy Hamilton
People who habitually suppress their emotions—inhibiting emotional expression—are more likely to experience severe nightmare-related distress, while those who use cognitive reappraisal, an adaptive reframing of emotional experiences, tend to report fewer nightmares. In a survey of university students, higher expressive suppression increased the odds of reporting more severe nightmare impairment by 14%. Higher cognitive reappraisal decreased the odds of more frequent nightmares by 4%. Age was not a significant factor. The findings suggest that emotion regulation styles are relevant to understanding vulnerability to nightmares, though causal mechanisms remain unclear.
Sleep
May 19, 2025
Garrett Baber, Nancy Hamilton, Matthew Gratton
ChatGPT's ratings of positive and negative affect in dream reports showed excellent agreement with participants' own ratings. In 136 dream reports, the model's positive affect ratings matched self-reports with an intra-class correlation coefficient of 0.844 and a mean absolute error of 1.778 on a 0–10 scale. For negative affect, agreement was similarly high, with an ICC of 0.857 and a mean absolute error of 1.681. Bland-Altman plots revealed no systematic bias. The findings suggest that large language models can reliably estimate emotional content in dream narratives, offering a scalable alternative to self-report for sleep and affective research.
Psychological Trauma
July 31, 2023
W. Youngren, N. Hamilton, Kristopher J Preacher et al.
Presleep cognitive arousal and longer sleep onset latency predict the occurrence of posttrauma nightmares in male veterans with mixed-trauma history. Physiological presleep arousal was also higher on nights when nightmares occurred. The findings replicate earlier work and suggest that therapies targeting presleep thoughts, time to fall asleep, and physical arousal may help treat posttrauma nightmares, though the small sample of 15 participants warrants cautious interpretation.