Bimodal Patterns of Locomotor Activity and Sleep in : A Model for Their Simulation.

Purpose: Two previously proposed modelling approaches to explain the bimodal pattern of activity and/or sleep in are based on 1) the concept of morning and evening oscillators underlying the peaks of activity in the morning and evening, respectively, and 2) the concept of two cycles of buildup and decay of sleep pressure, gated only by the circadian oscillator. Previously, we simulated 24-h alertness-sleepiness curves in humans using a model postulating the circadian modulation of the buildup and decay phases of two (wake and sleep) homeostatic processes. Here, we tested whether a similar model could be applied to simulate the bimodal 24-h rhythm of fly locomotor activity and sleep.

Methods: To obtain typical curves for the simulations, a sample of 4263 individual 24-h curves of locomotor activity and sleep were subjected to principal component analysis. It yielded three principal components, which explained more than 70% of the individual variations in these curves. We calculated the typical curves using scores on the 1, 2, and 3 principal components and simulated these curves and the sample-averaged curves.

Results: We found that these curves are always characterized by two peaks with varying sizes and timings. They can be fitted by proposing the variation of some of the parameters of the two homeostatic processes reflecting the 24-h rhythmicity of the drive for wake and the 12-h rhythmicity of the drive for sleep.

Conclusion: Postulation of two separate circadian oscillators is not necessary to explain the bimodal curves in .