The effect of sleep loss on working memory function in rested and sleep deprived patient with OSA

Introduction: Obstructive sleep apnoea (OSA) is rising in prevalenceand is linked with impaired performance on numerous cognitivedomains including working memory (WM) function. WM is importantfor efficient goal directed information processing, learning and taskperformance. WM function is impaired by sleep loss compromised inhealthy young subjects and may adversely affect safety during periods ofincreased sleep pressure (e.g. early morning). The effect of sleep loss onWM function in OSA patients is unclear. This study investigated WMfunction in OSA patients exposed to extended wakefulness andexplored clinical subjective sleepiness and sleep study metrics as WMpredictors.Methods: 58 patients with OSA underwent a baseline laboratory PSGfollowed by 28 hours of extended wakefulness (6:00am Saturday –10:00am Sunday). WM was measured 2 hourly over the 28 hours usingvisual 2- and 3-back tasks with completion time (CT) and accuracy(Acc) as main outcomes. These were averaged over day (2.00pm,4.00pm, 6.00pmpm) and night (12:00am, 2:00am, 5:00am). Correlations,regression and mixed models were performed to explore PSGvariables (sleep indices, AHI and oxygen indices) and subjective sleepiness(ESS) as correlates and predictors of WM.Results: There was a significant day to night decline in both N2 andN3-back CT and Acc performance (all p < 0.01). The effect size ofextended wakefulness on WM performance decline was very largeranging from 1.4–1.9 for Acc and (1.1–1.2) for CT. ESS andO2desaturation time <90% were significant predictors of day and nightN2-back CT R2 = 0.25, p < 0.01), while Acc was predicted by ESS and,and BMI (R2 = 0.25, p < 0.05). There were no significant predictors ofN3-back performance, the absolute day to night change in WM performance.Mixed models confirmed that sleepy patient (ESS ≥ 10) hadsignificantly impaired WM function compared with non-sleepy(ESS < 10) patients during both day and night time test, but this differencewas absent for the N3-back test.Conclusions: Sleep loss has a significant and large negative effect onWMfunction in OSA. Subjective sleepiness, hypoxemia and BMI appearto be best clinical predictors of N2-back, but as the task complexity isincreased, these factors were no longer predictive. The impact of WMimpairment on real life productivity and safety warrants furtherresearch.