Sleep deprivation (not sleep) facilitates short latency saccades: an exceptional example of unconscious procedural learning

Introduction: Saccades are visually triggered gaze shifts. Several cortical processing steps have to take place before the motor command is transferred to the brainstem and the eye muscles. The frequency distribution of saccadic reaction times shows at least two peaks: a narrow peak of short latency saccades (express saccades, 100–130ms) and a broader peak with regular saccades (about 180ms). For express saccades it is assumed that one processing step (the disengagement of attention) has already elapsed when the saccade target occurs. Furthermore, it is known that learning of express saccades builds up within days of training and then persists for a long time period. In this study we investigate how sleep and sleep deprivation effects the learning of express saccades.

Methods: Twenty healthy subjects performed visually triggered saccades in a gap-paradigm (test0). Then they performed a saccade training of about 1000 saccades. Thereafter, 10 subjects slept in a sleep laboratory environment, the other 10 subjects stayed awake. After 8 hours, the gap-paradigm was repeated in the morning (test1) and one day later (test2). Thus, one group performed a sleep deprivation/sleep sequence and the other a sleep/sleep sequence. After 4 weeks this procedure was repeated, but conditions were switched (the sleep group stayed awake and vice versa). In further control experiments we evaluated the effect of the training and the effect of sleep in the second night.

Results: Just after one single training session the sleep deprivation/sleep sequence significantly facilitated the learning process of express saccades. In contrast the sleep/sleep sequence showed hardly any effect. The learning effect persisted over the 4 week interval. Then the former sleep group had to stay awake and now showed the same learning effect for express saccades, while the former awake group now had to sleep without further benefit.

Conclusion: Learning of express saccades can be significantly enhanced by a sleep deprivation/sleep sequence. We assume, that sleep deprivation slightly reduces focused visual attention. This increases the probability for a state of disengaged attention, a prerequisite for the occurrence of express saccades. The slightly increased number of express saccades facilitates the learning process in Test2 after normal sleep.