BackgroundNew-onset postoperative atrial fibrillation is associated with adverse clinical outcomes in older adults. Poor preoperative sleep quality is a putative modifiable risk factor. The relationships between new-onset postoperative atrial fibrillation and preoperative sleep structure in older adults undergoing elective cardiac surgery were investigated at a single center. MethodsThis was a prespecified substudy within a prospective observational study of perioperative electroencephalographic markers (ClinicalTrials.gov; NCT03291626). We analyzed preoperative sleep recordings from 71 cardiac surgical patients aged ≥60 years without a prior history of atrial fibrillation. Overnight recordings were acquired using a consumer-grade headband and underwent manual sleep staging. Electroencephalographic slow wave activity (power in the 0.5-4 Hz frequency band) was computed in 1-minute intervals for non-rapid eye movement sleep stages. Associations between new-onset postoperative atrial fibrillation incidence and sleep measures were evaluated using univariate logistic regression models and multivariate logistic regression models including age and sex. ResultsNew-onset postoperative atrial fibrillation was present in 22 of 71 (31%) patients. A higher preoperative percentage of total sleep time in non-rapid eye movement stage 1 was associated with new-onset postoperative atrial fibrillation (median difference of 5.4%, P = .0002, Mann-Whitney U-test), independent of age and sex. No associations were observed between new-onset postoperative atrial fibrillation and other sleep metrics, including slow wave activity (all P > .05, Mann-Whitney U-test). ConclusionsExcess preoperative non-rapid eye movement stage 1 sleep, consistent with greater sleep fragmentation, is a potential modifiable target for mitigating new-onset postoperative atrial fibrillation risk in older adults undergoing elective cardiac surgery requiring cardiopulmonary bypass.