Dynamic Brain State Alterations in Narcolepsy: A Hidden Markov Model Approach to Thalamocortical Instability and...

Narcolepsy Type 1 (NT1) results from loss of hypothalamic orexin neurons, leading to unstable sleep-wake transitions. However, how this manifests as dynamic whole-brain functional state instability remains poorly characterized.

We applied Hidden Markov Modeling (HMM) to resting-state functional magnetic resonance imaging (fMRI) data from 30 patients with NT1 and 30 age- and sex-matched healthy controls to identify recurrent brain states, quantify their fractional occupancy (FO), and examine associations with clinical symptoms-specifically excessive daytime sleepiness (Epworth Sleepiness Scale, ESS) and hallucinations.

Five distinct dynamic brain states were identified. Compared to controls, NT1 patients showed significantly reduced FO in State 1 (thalamocortical arousal/sensory gating; p_FDR < 0.001) and elevated FO in State 4 (basal ganglia-limbic-sensorimotor integration; p_FDR < 0.001) and State 5 (reward-introspection; p_FDR = 0.036). Notably, within the NT1 group, patients with hallucinations exhibited higher FO in State 1 than those without (p_FDR = 0.042), suggesting aberrant recruitment of this state during sleep-wake transitions. Additionally, State 4 FO showed a moderate positive correlation with ESS scores (Spearman's ρ = 0.38, p_FDR = 0.078).

NT1 is associated with reduced stability of a thalamocortical alertness state and increased expression of REM-like limbic-subcortical network configurations. State 1 occupancy is relatively elevated in patients with hallucinations, while State 4 shows a positive association with excessive daytime sleepiness and features of REM dissociation. These findings support a dynamic, whole-brain systems-level framework for understanding symptom heterogeneity in NT1.