Nightmares are distressing and linked to psychiatric problems, but not everyone with high negative affect experiences them. This survey of 846 UK adults found that, beyond negative affect, higher levels of worry, depersonalisation, hallucinatory experiences, paranoia, and longer sleep duration were associated with nightmare occurrence (odds ratios 1.25–1.45). Nightmare severity was associated with worry, depersonalisation, hallucinatory experiences, and paranoia. Alcohol use and physical activity were not associated with nightmares. The findings suggest several potential predictors of nightmares that need testing in future causal studies.
Sleep dysfunction and hallucinatory experiences are bidirectionally related, with the strongest path leading from poor sleep to hallucinations. This narrative review examines mechanisms across four levels: phenomenology, psychology, neural networks, and neurophysiology. Stress is a well-supported mediator, with sleep manipulation studies in non-clinical populations showing its role. Sleep loss also affects nervous system inflammation, altering brain connectivity involved in hallucinations. Lived-experience accounts reveal three novel mechanisms—source monitoring, mental resilience, and reasoning skills—that are impacted by sleep loss, though the full causal chain requires further testing. Priorities include testing stress as a mediator in clinical populations and experimentally examining these novel mediators.
Imagery rehearsal therapy, a nightmare treatment, was acceptable and feasible for people who also have psychotic symptoms. In a small sample of six participants with frequent distressing nightmares and psychosis, five completed four to six sessions of imagery rehearsal. Nightmare-related distress, vividness, and intensity showed descriptive improvements, and participants gave positive feedback. Nightmare frequency did not decrease, but participants reported a change in their emotional response to nightmares. A larger controlled trial is needed to test the therapy's efficacy for this population.