Quantitative sleep EEG and polysomnographic predictors of driving simulator performance in obstructive sleep apnea.

Objectives: To improve identification of obstructive sleep apnea (OSA) patients at risk of driving impairment, this study explored predictors of driving performance impairment in untreated OSA patients using clinical PSG metrics, sleepiness questionnaires and quantitative EEG markers from routine sleep studies.

Methods: Seventy-six OSA patients completed sleepiness questionnaires and driving simulator tests in the evening of their diagnostic sleep study. All sleep EEGs were subjected to quantitative power spectral analysis. Correlation and multivariate linear regression were used to identify the strongest predictors of driving simulator performance.

Results: Absolute EEG spectral power across all frequencies (0.5-32 Hz) throughout the entire sleep period and separately in REM and NREM sleep, (r range 0.239-0.473, all p<0.05), as well as sleep onset latency (r=0.273, p<0.017) positively correlated with driving simulator steering deviation. Regression models revealed that amongst clinical and qEEG variables, the significant predictors of worse steering deviation were greater total EEG power during NREM and REM sleep, greater beta EEG power in NREM and greater delta EEG power in REM (range of variance explained 5-17%, t range 2.29-4.0, all p<0.05) and sleep onset latency (range of variance explained 4-9%, t range 2.15-2.5, all p<0.05).

Conclusions: In OSA patients, increased EEG power, especially in the faster frequency (beta) range during NREM sleep and slower frequency (delta) range in REM sleep were associated with worse driving performance, while no relationships were observed with clinical metrics e.g. apnea, arousal or oxygen indices.

Significance: Quantitative EEG analysis in OSA may provide useful markers of driving impairment risk. Future studies are necessary to confirm these findings and assess the clinical significance of quantitative EEG as predictors of driving impairment in OSA.