Abstract
BACKGROUND: Actigraphy has received increasing attention in classifying rest-activity cycles. However, in patients with disorders of consciousness (DOC), actigraphy data may be considerably confounded by passive movements, such as nursing activities and therapies. Consequently, this study verified whether circadian rhythmicity is (still) visible in actigraphy data from patients with DOC after correcting for passive movements. METHODS: Wrist actigraphy was recorded over 7-8 consecutive days in patients with DOC (diagnosed with unresponsive wakefulness syndrome [UWS; n = 19] and [exit] minimally conscious state [MCS/EMCS; n = 11]). The presence and actions of clinical and research staff as well as visitors were indicated using a tablet in the patient's room. Following removal and interpolation of passive movements, non-parametric rank-based tests were computed to identify differences between circadian parameters of uncorrected and corrected actigraphy data. RESULTS: Uncorrected actigraphy data overestimated the interdaily stability and intradaily variability of patients' activity and underestimated the deviation from a circadian 24-h rhythm. Only 5/30 (17%) patients deviated more than 1 h from 24 h in the uncorrected data, whereas this was the case for 17/30 (57%) patients in the corrected data. When contrasting diagnoses based on the corrected dataset, stronger circadian rhythms and higher activity levels were observed in MCS/EMCS as compared to UWS patients. Day-to-night differences in activity were evident for both patient groups. CONCLUSION: Our findings indicate that uncorrected actigraphy data overestimates the circadian rhythmicity of patients' activity, as nursing activities, therapies, and visits by relatives follow a circadian pattern itself. Therefore, we suggest correcting actigraphy data from patients with reduced mobility.