Clinical predictors of working memory performance in obstructive sleep apnea patients before and during extended wakefulness

Study Objectives: Extended wakefulness (EW) and obstructive sleep apnea (OSA) impair working memory (WM), but their combined effects are unclear.This study examined the impact of EW on WM function in OSA patients and identified clinical predictors of WM impairment. Methods: Following polysomnography (PSG), 56 OSA patients (mean +/- SD, age 49.5 +/- 8.9, apnea hypopnea index 38.13 +/- 25.0) completed WM 2-back performance tasks 10 times over 24 h of wakefulness to assess average accuracy and completion times measured after 6-12 h awake (baseline) compared to 18-24 h awake (EW). Hierarchical cluster analysis classified participants with poorer versus better WM performance at baseline and during EW. Clinical predictors of performance were examined via regression and receiver operator characteristic (ROC) analyses. Results: WM performance decreased following EW and showed consistent correlations with age, Epworth Sleepiness Score (ESS), total sleep time, and hypoxemia (O-2 nadir and mean O-2 desaturation) at baseline and with EW (all p < .01). O-2 nadir and age were significant independent predictors of performance at baseline (adjusted R-2 = 0.30, p < .01), while O-2 nadir and ESS were predictors of WM following EW (adjusted R-2 = 0.38, p < .001). ROC analysis demonstrated high sensitivity and specificity of models to predict poorer versus better performing participants at baseline (83% and 69%) and during EW (84% and 74%). Conclusions: O(2 )nadir, age, and sleepiness show prognostic value for predicting WM impairment in both rested and sleep-deprived OSA patients and may guide clinicians in identifying patients most at risk of impaired WM under both rested and heightened sleep pressure conditions.