Verbal memory is linked to average oxygen saturation during sleep, not the apnea-hypopnea index nor novel hypoxic...

The apnea-hypopnea index (AHI) is the current diagnostic parameter for diagnosing and estimating the severity of obstructive sleep apnea (OSA). It is, however, poorly associated with the main clinical symptom of OSA, excessive daytime sleepiness, and with the often-seen cognitive decline among OSA patients. To better evaluate OSA severity, novel hypoxic load parameters have been introduced that consider the duration and depth of oxygen saturation drops associated with apneas or hypopneas. The aim of this paper was to compare novel hypoxic load parameters and traditional OSA parameters to verbal memory and executive function in OSA patients.

A total of 207 adults completed a one-night polysomnography at sleep laboratory and two neuropsychological assessments, the Rey Auditory Verbal Learning Test and Stroop test.

Simple linear regression analyses were used to evaluate independent associations between each OSA parameter and cognitive performance. Associations were found between immediate recall and arousal index, hypoxia <90 %, average SpO₂ during sleep, and DesSev100+RevSev100. Total recall was associated with all OSA parameters, and no associations were found with the Stroop test. Subsequently, sex, age, and education were included as covariates in multiple linear regression analyses for each OSA parameter and cognitive performance. The main findings of the study were that average SpO₂ during sleep was a significant predictor of total recall (p < .007, β = -.188) with the regression model explaining 21.2 % of performance variation. Average SpO₂ during sleep was also a significant predictor of immediate recall (p < .022, β = -.171) with the regression model explaining 11.4 % of performance variation. Neither traditional OSA parameters nor novel hypoxic load parameters predicted cognitive performance after adjustment for sex, age, and education.

The findings validate that the AHI is not an effective indicator of cognitive performance in OSA and suggest that average oxygen saturation during sleep may be the strongest PSG predictor of cognitive decline seen in OSA. The results also underline the importance of considering age when choosing neurocognitive tests, the importance of including more than one test for each cognitive domain as most tests are not pure measures of a single cognitive factor, and the importance of including tests that cover all cognitive domains as OSA is likely to have diffuse cognitive effects.