Measures of sleep fragmentation, including N1 percentage, stage shift index, and wake after sleep onset, along with daytime Maintenance of Wakefulness Test (MWT) latencies reliably differentiate narcolepsy type 1 (NT1) from type 2 (NT2), according to a test-retest reliability study published July 7 in Sleep.
While quantitative EEG features showed the highest reproducibility overall, they rarely distinguished between the two diagnoses. The findings provide clinicians with a practical shortlist of objective measures that are both reliable and diagnostically informative.
What they found
Researchers analyzed placebo-arm data from two clinical trials, 17 drug-free NT1 patients (NCT05687903) and 19 drug-free NT2 patients (NCT05687916), who each completed three clinic visits four weeks apart. From nocturnal polysomnography and next-day MWT recordings, the team extracted 440 features spanning quantitative EEG, hypnogram-based sleep architecture, and daytime sleep latency measures.
Reliability of key feature categories:
| Feature Category | Top Features Reliability | Notes |
|—|—|—|
| Quantitative EEG | r > 0.88 | Highest overall, but rarely distinguished NT1 vs NT2 |
| Hypnogram (sleep architecture) | r = 0.68–0.78 | Moderate-to-high reliability |
| MWT sleep onset latencies (2pm, 4pm) | r = 0.87–0.89 (NT1) | Good in NT1 only; less reliable in NT2 |
| Sleep fragmentation (N1%, shift index, WASO) | r = 0.54–0.75 | Reliable AND differentiated between diagnoses |
Diagnostic differences (NT1 vs NT2):
- N1 percentage, stage shift index, and WASO were all significantly higher in NT1 (p < 0.05)
- MWT sleep onset latencies at 2 p.m. and 4 p.m. were shorter in NT1 (p = 0.037 and 0.010, respectively)
- Hypnodensity Wake-REM mixtures, a novel metric not yet used clinically, showed strong reliability (r = 0.83) and were significantly higher in NT1 (p = 0.001)
NT2 patients exhibited substantially greater variability across all three visits compared to NT1, suggesting that the disorder’s nighttime and daytime manifestations fluctuate more over time.
Why it matters
Distinguishing NT1 from NT2 has therapeutic implications, NT1 involves hypocretin deficiency and typically requires different management strategies, including sodium oxybate and pitolisant, whereas NT2, a less well-characterized condition, lacks a clear biomarker. The study identifies a compact set of polysomnography-derived metrics that clinicians can use immediately, without specialized analysis, to support differential diagnosis.
The hypnodensity Wake-REM finding is particularly notable: it identifies a spectral signature of REM sleep intrusion into wakefulness that is both highly reproducible and disease-specific, potentially serving as a future diagnostic feature.
Limits
The sample sizes are modest (17 NT1, 19 NT2), and the data come from clinical trial placebo arms, which may not fully represent the general narcolepsy population as trial participants undergo rigorous screening. The study also cannot address whether these features predict treatment response.
Bottom line
Sleep fragmentation indices and MWT latencies, already part of standard sleep center protocols, reliably differentiate NT1 from NT2. The addition of hypnodensity analysis may further sharpen diagnostic accuracy once clinical workflows adopt it.
Source
Schlafly E, et al. “Nocturnal features and daytime characteristics in narcolepsy: Reliability and diagnostic relevance for NT1 vs NT2.” Sleep. 2026 Jul 7;:zsag180. DOI: 10.1093/sleep/zsag180