Sleep fragmentation and working memory in healthy adults

Introduction: Sleep is essential for performing cognitive function in humans. We have hypothesized that sleep fragmentation compared to sleep efficiency may have a negative impact on the working memory. Material and Methods: Twenty-eight healthy adults (18 males and 10 females; mean age 27.8 +/- 15.5 years) were enrolled in this study. We measured the total sleep time (TST), sleep efficiency, %stage wakefulness (W), %stage rapid eye movement (REM), %stage N1, %stage N2, %stage N3, wake after sleep onset (WASO), and arousal index using polysomnography. Working memory, executive function, and sustained attention of three domains of cognitive function were evaluated with the number of back task (N-back task), Wisconsin card sorting test (WCST), and continuous performance test-identical pairs (CPT-IP), respectively. Results: The percentage of correct answers on the 2-back task was significantly correlated with %stage REM, %stage N1, and %stage N2 (%stage REM: r=0.505, p=0.006; %stage N1: r=-0.637, p<0.001; %stage N2: r=0.670, p<0.001), and multiple regression analysis including the stepwise forward selection method revealed that %stage N2 was the most significant factor (%stage N2: beta=0.670, p<0.001). The percentage of correct answers on the 2-back task was also significantly correlated with TST, sleep efficiency, WASO, and arousal index (TST: r=0.492, p=0.008; sleep efficiency: r=0.622, p<0.001; WASO: r=0.721, p<0.001; arousal index: r=-0.656, p<0.001), and WASO was the significant factor (beta=-2.086, p=0.007). The WCST category achievement and CPT-IP d-prime score were correlated with none of the sleep variables. Conclusion: Increased WASO and a decrease in %stage N2 were associated with worse working memory.