Nightmare disorder has long been considered a REM parasomnia, but new research from King’s College London reveals that its signature may lie in the NREM sleep microarchitecture instead.
In a study of 26 adults with nightmare disorder and 32 controls, published in NPJ Biological Timing and Sleep, Istvan Papp and colleagues found that people with frequent nightmares have reduced coupling between slow oscillations and sleep spindles during NREM sleep, a brain-rhythm coordination critical for memory consolidation and neural repair.
What they found
Using overnight polysomnography with a harmonised event-based EEG pipeline, the researchers focused on the temporal coordination of two hallmark NREM rhythms: slow oscillations (slow waves that sweep across the cortex) and sleep spindles (brief bursts of oscillatory activity).
Their key finding was a reduction in what they term slow-oscillation-dominant spindle coupling in frontal NREM sleep. The coupling dominance score, which measures how strongly slow oscillations, rather than delta waves, drive spindle timing, was significantly lower in the nightmare group:
- Nightmare disorder: 0.137 ± 0.083
- Controls: 0.238 ± 0.096
- p = 6.8 × 10⁻⁵
Delta-spindle coupling, by contrast, was preserved between the groups.
A secondary finding was reduced K-complex density, the rate of these sleep-specific brainwave events, across the first six hours from sleep onset in nightmare disorder (0.463 ± 0.345 versus 0.723 ± 0.475 events per minute of N2/N3 sleep; p = 0.019).
Why it matters
The traditional view holds that nightmare disorder is primarily a REM phenomenon, since nightmares themselves are REM-based. These findings challenge that assumption by showing that NREM sleep microarchitecture is fundamentally altered in people with nightmare disorder.
The specific deficit in slow-oscillation-spindle coupling is notable because this coupling is believed to support systems consolidation of memories overnight. Disrupted coupling could mean that people with nightmare disorder do not just experience disturbing dreams, their basic sleep-dependent brain maintenance may also be compromised.
Limits
The study was retrospective and relatively small (26 cases, 32 controls). The peak-timing differences for K-complexes were not statistically robust. The authors themselves note that prospective confirmation in larger, harmonised, clinically phenotyped cohorts is needed before these findings can be considered definitive.
Bottom line
Nightmare disorder is not just a REM sleep problem. People with frequent nightmares show altered NREM sleep microarchitecture, specifically reduced slow-oscillation-driven spindle timing, suggesting the condition involves a broader disruption of basic sleep brain rhythms than previously recognized.
Source
Papp I, et al. Nightmare disorder shows reduced slow oscillation-dominant spindle coupling in NREM sleep. NPJ Biol Timing Sleep. 2026 Jun 22;3(1):28. doi:10.1038/s44323-026-00094-0. PMID: 42331979.