Long COVID Leaves Distinct Fingerprints in Brain Activity During Sleep
New EEG research reveals measurable sleep microstructure abnormalities in Long COVID patients, offering clues to why rest feels unrefreshing.
Summary
Researchers at Beth Israel Deaconess and Massachusetts General Hospital used detailed overnight brain wave recordings to compare sleep patterns in Long COVID patients, healthy controls, and people with ME/CFS. They found that Long COVID patients showed abnormal slow oscillations during waking and REM sleep, disrupted sleep spindle behavior, and altered coupling between brain wave patterns that normally support restorative sleep. These changes were distinct but partially overlapping with ME/CFS. The findings provide a biological explanation for the persistent fatigue and unrefreshing sleep that Long COVID patients commonly report, and suggest that measurable brain activity changes — not just subjective complaints — underlie these symptoms. This opens the door to objective diagnostic markers and potentially targeted treatments for post-infectious sleep dysfunction.
Detailed Summary
Millions of people recovering from COVID-19 report persistent fatigue and sleep that never feels restorative, yet the biological mechanisms behind these symptoms have remained poorly understood. This study takes an important step toward explaining why, using detailed brain wave analysis during sleep.
Researchers recruited 28 Long COVID patients and 28 age- and sex-matched healthy controls for overnight polysomnography — comprehensive in-lab sleep studies — at a single research center. An additional 19 ME/CFS patients from a clinical center were included for comparison. The team analyzed fine-grained EEG features including spectral band powers, sleep spindles, slow oscillations, spindle-slow oscillation coupling, brain age index, alpha-delta patterns, and infraslow oscillations.
The results were striking. Long COVID patients showed elevated slow oscillation power during waking before sleep and during REM sleep. Their sleep spindles exhibited a faster frequency drop within each spindle and shorter slow oscillation peaks in frontal brain regions. Critically, spindle-slow oscillation coupling — a mechanism thought to support memory consolidation and restorative sleep — was disrupted, occurring earlier than normal relative to the slow oscillation trough. This early coupling was directly associated with worse subjective sleep quality.
ME/CFS shared some of these abnormalities but also showed distinct features: higher uncoupled slow spindle densities, more alpha-delta intrusions early in the night, and elevated infraslow oscillation power — patterns previously linked to non-restorative sleep in other conditions.
These findings matter because they provide objective, measurable evidence that Long COVID and ME/CFS alter brain activity during sleep in specific, potentially pathological ways. This could eventually support development of EEG-based biomarkers for diagnosis and treatment monitoring.
Caveats include the small sample sizes, the cross-sectional design, and the fact that ME/CFS data came from a separate clinical center rather than the same research protocol, limiting direct comparisons.
Key Findings
- Long COVID patients show abnormal slow oscillation power during waking and REM sleep compared to healthy controls.
- Sleep spindle-slow oscillation coupling is disrupted in Long COVID, with earlier coupling linked to worse sleep quality.
- Long COVID spindles show faster frequency drop and shorter slow oscillation peaks in frontal brain regions.
- ME/CFS shares some Long COVID EEG abnormalities but also shows distinct alpha-delta intrusions and elevated infraslow oscillations.
- Objective EEG microstructure changes may explain the unrefreshing, non-restorative sleep reported by Long COVID patients.
Methodology
The study used overnight in-lab polysomnography in 28 Long COVID patients and 28 age-sex-matched healthy controls from a single research center, plus 19 ME/CFS patients from a separate clinical center. EEG microstructure analysis included spectral band powers, sleep spindles, slow oscillations, spindle-SO coupling, brain age index, alpha-delta patterns, and infraslow oscillations. The study was cross-sectional and observational.
Study Limitations
Sample sizes were small (28 per group for LC vs HC), limiting statistical power and generalizability. ME/CFS data came from a separate clinical center with a different protocol, making direct comparisons less rigorous. The summary is based on the abstract only, as the full text was not available.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.