Rapid eye movements are reduced in blind individuals.

There is ongoing controversy regarding the role of rapid eye movements (EMs) during rapid eye movement (REM) sleep. One prevailing hypothesis is that EMs during REM sleep are indicative of the presence of visual imagery in dreams. We tested the validity of this hypothesis by measuring EMs in blind subjects and correlating these with visual dream content. Eleven blind subjects, of whom five were congenitally blind (CB) and six late blind (LB), and 11 matched sighted control (SC) subjects participated in this study. All participants underwent full-night polysomnography (PSG) recordings that were staged manually following American Academy of Sleep Medicine (AASM) criteria. Nocturnal EMs were detected automatically using a validated EM detector, and EM activity was represented as "EM coverage" computed as percentage of time with EM in each sleep stage. Frequency of sensory dream elements was measured in dream recall questionnaires over a 30-day period. Both blind groups showed less EM coverage during wakefulness, N1, N2 and REM sleep than did controls. CB and LB subjects did not differ in EM activity. Validation of the detector applied to blind subjects revealed an overall accuracy of 95.6 ± 3.6%. Analysis of dream reports revealed that LB subjects reported significantly more visual dream elements than did CB. Although no specific mechanisms can be revealed in the current study, the quasi absence of nocturnal EMs in LB subjects despite preserved visual dream content does not support the visual scanning of dreams hypothesis. Specifically, results suggest a dissociation between EMs and visual dream content in blind individuals.