Night-to-night variability of sleep latency significantly predicts the magnitude of subsequent change in sleep latency during placebo administration

Objective: To investigate the association between subjective sleep latency (sSL) fluctuation during the lead-in period and the placebo response in the subsequent double-blind period. Methods: The current study is a secondary analysis of data from the placebo arm (380 patients) of a double-blind study in outpatients with primary insomnia. Results: Higher fluctuation of lead-in sSL was associated with a greater decrease in sSL in the subsequent weeks. Multivariate analysis suggested that a wider standard deviation for daily sSL and a higher weekly mean sSL during the lead-in period were independent predictors of greater improvement in mean sSL during the subsequent weeks of placebo treatment. Likewise, a wider standard deviation and lower mean of subjective total sleep time (sTST) during the lead-in period were independent predictors of greater improvement in mean sTST in the subsequent weeks, but predictability of change in sTST with these parameters appears lower than those for sSL (R-2 = 0.13 and 0.44, respectively, in Week 2). Conclusions: The importance of night-to-night variability of sSL in sleep assessment is highlighted. Excluding patients with high variability of lead-in sSL should be considered in clinical studies evaluating an effect on sleep onset, especially in the early stages of clinical development. (C) 2011 Elsevier B.V. All rights reserved.