Dimensional analysis of actigraphic derived sleep data

Nonlinear dimensional analyses can be a useful tool in understanding the underlying behavior of dynamical systems, including biological systems. Many biological functions can be modeled as chaotic processes, including sleep. Sleep data can be obtained from several methods, such as electroencephalograms, polysomnography, and actigraph. Actigraphy, because of its low level of invasiveness, is an increasingly popular method of obtaining sleep data. This study analyzed actigraphy data with nonlinear dimensional analyses to determine if such analytic methods would be useful in sleep studies. Participants wore actigraphs on their wrists, which recorded movement for several days. Several sleep quality variables, such as movement during sleep and total sleep time, were derived from these sleep data. These variables were used to determine whether the quality of sleep was good or poor. Lagged phase space plots were graphed and nonlinear parameters for the fractal dimension and the correlation dimension were computed for each participant. Descriptive and inferential statistics were performed to determine if the nonlinear parameters showed significant differences with respect to sleep quality.