Identifying Sleep Biomarkers to Evaluate Cognition in HIV

Area of Science:

• Neuroscience
• Infectious Diseases
• Sleep Medicine

Background:

• Sleep disturbance and cognitive impairment are prevalent in individuals with HIV, even with effective antiretroviral therapy.
• Understanding the link between sleep disruption and cognitive function in HIV+ individuals is crucial for improved treatment.
• Biomarkers derived from sleep assessments offer a pathway to elucidate these mechanisms.

Purpose of the Study:

• To identify sleep-assessment-based biomarkers for distinguishing cognitive impairment in HIV+ individuals.
• To develop a risk score combining polysomnography (PSG) features to predict cognitive status.
• To explore the utility of EEG, EMG, and EOG signals in assessing cognitive function.

Main Methods:

• Utilized polysomnography (PSG) data including EEG, EMG, and EOG signals from 18 HIV+ Black males.
• Employed a generalized linear model (GLM) to compute a risk score integrating PSG features.
• Applied the least absolute shrinkage and selection operator (LASSO) for optimal feature selection.

Main Results:

• A significant risk separation (P < .001) was achieved between cognitively normal and impaired HIV+ groups.
• The optimal predictive features were exclusively EEG-derived and specific to sleep stages.
• The developed risk score effectively distinguished between cognitive states.

Conclusions:

• Sleep-based EEG features show promise as diagnostic and prognostic biomarkers for cognitive function in HIV+ individuals.
• This study highlights the potential of polysomnography-derived metrics in managing cognitive health in HIV.
• Further research can validate these sleep-based biomarkers for clinical application in HIV care.