Meta-analysis of the antidepressant response to sleep deprivation and its correlates: towards a better antidepressant therapy

Unlike antidepressant drugs, which typically require several weeks to produce an antidepressant response, sleep deprivation produces a response literally overnight. Quantification (meta-analysis) of 166 articles, including data from a total of 3951 depressed patients, reveals that consistently half of all depressed patients are responders to a night of sleep deprivation, with the degree of response shown by these responders being on average a 55% decrease in depression levels. While the level of this response depends upon both when the sleep deprivation occurs in the night and when response measurements are taken, no experimental treatment has yet been found to further enhance its response. The practicality of sleep deprivation as an antidepressant treatment has so far been limited by the fact that the majority of responders to sleep deprivation normally relapse by the day following a night of recovery sleep. However, there is some evidence that this relapse can be prevented or delayed, especially by depletion of the serotonergic system. The strength of reported correlates of response to sleep deprivation and of its relapse were examined and the nature of the most powerful correlates of response was found to depend upon their timing: correlates measured before sleep deprivation (thus related to the susceptibility to response) show between-subjects differences while correlates dependent upon measurements taken after a night of sleep deprivation (thus related to the response itself) show only within-subject changes from before to after sleep deprivation. Since whether a patient is a responder to one night of sleep deprivation is unrelated to whether the same patient will be a responder to any other night of sleep deprivation, it is hypothesized that the activity levels of some of these predictor variables may also change across time in relation to the susceptibility to response. The discovery of such susceptibility-state markers and their temporal order could help shed light on the mechanism of susceptibility to response and thus offer new ways of improving current antidepressant treatments