0306 MULTISCALE ENTROPY OF NOCTURNAL HEART RATE IN CONJUNCTION WITH POLYSOMNOGRAPHY: A NOVEL APPROACH TO UNDERSTAND PHYSIOLOGICAL COMPLEXITY IN INSOMNIA

Abstract

Insomnia is the most prevalent sleep disorder and one of the most common complaints in primary care. Despite this, the pathophysiology underpinning insomnia complaints are still poorly understood. There have been numerous attempts to investigate the possibility of nocturnal heart rate (HR) and its variability as physiological hallmarks of the hyperarousal experienced in insomnia, with mixed results. We employed Multiple Scale Entropy analysis (MSE) of HR, a known hallmark of complexity in cardiac system controls, in conjunction with home polysomnography (PSG) over two nights, to investigate patterns of HR complexity across sleep stages in 23 students with subthreshold insomnia (ISI>9) and 20 good sleepers (18–30 years, gender matched). MSE was computed over multiple temporal intervals (scale factors) derived from uninterrupted periods of HR. Linear mixed models were fit for each sleep stage using scale factors as the outcome measures. We found a significant group X sleep stage interaction in the prediction of nocturnal HR (F(41343, 5) = 159.7, p<0.0001). Entropy-based trends revealed a decrease in entropy with increasing time series length in insomnia, while entropy increased with time series length in good sleepers. Significant differences in entropy were observed at scale 18 for Stage 2 sleep and in scales 9, 10,12,13,14 and 18 in Stage 3 (p= 0.023) and at scales 10 and 12 (p=0.011) in REM sleep between groups. Our results indicate that the biological complexity of the HR signal significantly differs between the groups as coarse-grained time series become progressively more regular and less complex for the insomnia group across different stages of sleep. This could be a signal of degradation of control mechanisms over long time scales. MSE analysis of HR data, in conjunction with PSG, appears a promising avenue for distinguishing biological complexity in insomnia populations. The study is supported by both the Medical Research Council (studentship to JC) and a Wellcome Trust Strategic Award (098461/Z/12/Z) to the Oxford Sleep and Circadian Neuroscience Institute (SCNi).