Earlier studies reported that subjective sleep quality correlates with objective polysomnographically measured sleep initiation and continuity, but less with sleep architecture.

This study aimed to investigate relations between subjective and objective sleep in normals (N), insomnia comorbid with generalized anxiety disorder (G) and apnea (A).

One hundred and seventy-seven normals (50.9±19.6a), 100 insomniac G patients (37.9±10.6a) and 51 A patients (51.3±9.7a) completed the self-rating scale for sleep and awakening quality (Saletu et al. 1997) regarding two polysomnographic nights analyzed by the Somnolyzer (Anderer et al. 2005). Correlations were calculated based on changes between the first (adaptation) and second polysomnographic night to diminish inter-individual variances of sleep perception.

In N, subjective sleep quality (S-QUA) demonstrated correlations (p < 0.01) with sleep efficiency (EFF), wake after sleep onset (WASO), S2, S1%, REM, S1, frequency of awakenings (FW), latency to continuous sleep (L-CONT), sleep latency (S-LAT) and slow-wave sleep, while awakening quality (A-QUA) showed weak (p < 0.05) correlations with EFF and WASO. Somatic complaints (S-COM) correlated (p < 0.05) with WASO and REM. In G, correlations (p < 0.01) were obtained between S-QUA and EFF, WASO, S2, L-CONT and S-LAT (p < 0.05), while A-QUA correlated with S2, WASO and EFF. In A, S-QUA correlated (p < 0.01) with EFF, S2, S1%, S2%, L-CONT, WASO and less (p < 0.05) with S-LAT and S1. A-QUA correlated with S2, S2% (p < 0.01), L-CONT and EFF (p < 0.05). S-COM correlated with S-CONT (p < 0.01) and S-LAT (p < 0.05).

EFF, WASO, S2 and less S-LAT determine good S-QUA in all groups.