REM sleep fragmentation and arousability distinguish chronic insomnia subtypes and reflect subjective-objective sleep discrepancy.

STUDY

OBJECTIVES

The objective was to characterize rapid eye movement (REM) sleep structure in chronic insomnia patients with objective polysomnography (PSG)-defined macrostructural abnormalities (CIN-AP) or without such abnormalities (CIN-NP), and to determine how these subtypes differ from each other and from good sleepers (GS).

METHODS

We analyzed 1049 participants: 141 GS, 214 CIN-NP, and 694 CIN-AP. CIN was defined by ICSD-3 criteria. Automated algorithms quantified sleep stages, arousals, and REM features including percentage and duration (total, phasic, tonic), number of arousals and awakenings per hour, and rapid eye movements per hour (REM density). Secondary analyses examined arousal dynamics via odds ratio product (ORP) and awakening duration distributions. Group differences were assessed with multinomial logistic regression, and stepwise regression identified independent predictors of CIN-AP versus CIN-NP.

RESULTS

Both CIN subtypes showed more REM arousals and awakenings per hour, higher REM density, and greater phasic REM proportion compared with GS. Notably, these REM abnormalities were present in CIN-NP despite the absence of conventional PSG-macrostructural abnormalities. Compared with CIN-NP, CIN-AP exhibited longer REM awakening durations but similar REM arousal and awakening indices. Independent predictors distinguishing CIN-AP from CIN-NP included male sex, older age, lower REM duration, higher REM percentage, increased phasic REM, and longer REM awakenings. The predictive model achieved an AUC of 0.752 (65.6% sensitivity, 72.3% specificity); inclusion of ORP and awakening duration variables yielded only marginal improvement (AUC 0.755, 68.4% sensitivity, 69.7% specificity).

CONCLUSIONS

REM instability distinguishes chronic insomnia subtypes and is detectable even in patients lacking conventional PSG abnormalities. These findings suggest that REM microstructure captures clinically relevant aspects of insomnia not reflected by standard sleep architecture measures and may provide objective physiological markers for phenotype-specific diagnosis and treatment.