Feed

Sleep disturbances are common in chronic pain populations and often assessed with actigraphy, which estimates sleep indirectly from movement. Frequent nocturnal movement or extended periods of motionless wakefulness may decrease the accuracy of movement-based sleep detection in this population. Wearable electroencephalography (EEG) offers an alternative to actigraphy by directly capturing neurophysiological sleep signals in home settings. We conducted a 7-day home-based study in adults with chronic musculoskeletal pain (n=21; 71% female; 76% White), where participants wore a wrist actigraph and EEG headband and completed sleep diaries. Limits of agreement (LoA) were assessed using Bland-Altman analyses. Relative to wearable EEG, actigraphy showed substantial disagreement for all sleep variables with proportional bias: TST (mean bias=1.79 min; LoA=-122.02 to 125.60 min), sleep efficiency (mean bias= -0.90%; LoA=-22.01 to 20.20%), wake after sleep onset (mean bias=22.86 min; LoA=-75.23 to 120.95 min), and awakenings (mean bias=-5.70 events; LoA=-31.87 to 20.47 events). Epoch-by-epoch analyses demonstrated high sensitivity (92.8%) but low specificity (49.7%), with agreement of 87.6% and Cohen's κ = 0.42, and revealed substantial within- and between-person variability. Wide LoA, proportional bias, and weak epoch-level agreement suggest that actigraphy should be interpreted cautiously when used as a surrogate for neurophysiologic sleep assessment in chronic pain. PERSPECTIVE: This 7-day home-based study highlights the need for context-sensitive sleep assessment in chronic pain: movement-based actigraphy may be useful in some settings but may be a poor surrogate for wearable EEG when accurate wakefulness, sleep fragmentation, or night-to-night variability estimates are needed.