Abstract
INTRODUCTION: Patients with implantable cardioverter defibrillators (ICDs) frequently experience sleep disruption. Prior work shows associations between objective (actigraphic) sleep and cognition in these patients, but whether pain affects associations between measures of sleep fragmentation (eg, sleep efficiency, SE) and cognition is unknown. The present study examined independent and interactive associations between objective SE and pain on cognitive performance in patients with ICDs. METHODS: A total of 37 patients with ICDs (M(age) = 60.0, SD = 12.4) and self-reported sleep disturbance completed 14 days of actigraphy. Average SE was computed [(average total sleep time/average time in bed) × 100%]. Patients completed the Short Form 36 Health Survey pain section, and computerized tasks measuring executive functioning (letter series, N-Back task), sustained attention/processing speed (symbol digit modalities test, SDMT), and simple reaction time. Multiple linear regressions examined whether SE independently predicted or interacted with pain ratings to predict cognitive performance. RESULTS: SE interacted with pain to predict SDMT performance, accounting for 12% unique variance. In patients reporting worse pain, higher SE was associated with better SDMT performance. Similar patterns of association on SDMT were not observed in patients with average or low pain. SE and pain ratings did not independently predict SDMT performance. Performance on other cognitive tasks was not associated with any predictors. CONCLUSION: Better sleep efficiency may play an important role in improving sustained attention/processing speed in patients with ICDs and perceived severe pain. Future research should examine whether interventions aimed at improving sleep fragmentation provide benefit to lower order cognition, particularly in patients with worse pain.