Detecting isolated REM sleep behavior disorder at home using a lower-back wearable sensor.

Area of Science:

• Neurology
• Sleep Medicine
• Biomedical Engineering

Background:

• Isolated REM Sleep Behavior Disorder (iRBD) predicts synucleinopathies.
• Current diagnosis relies on resource-intensive polysomnography (vPSG) in sleep labs.
• Home-based diagnostics are needed due to limited vPSG access and sleep habit variations.

Purpose of the Study:

• Evaluate feasibility of home-based iRBD identification using lumbar wearable sensors.
• Assess night-to-night variability in sensor data.
• Explore potential for a staged screening approach.

Main Methods:

• Seventy-three participants (15 iRBD, 58 controls) underwent vPSG.
• Participants wore a lumbar inertial measurement unit for six nights at home.
• Machine learning models analyzed mobility features for iRBD classification.

Main Results:

• iRBD participants exhibited distinct nocturnal mobility patterns compared to controls.
• Machine learning models achieved high sensitivity and moderate specificity for iRBD detection.
• Diagnostic performance improved with data from up to five nights, then plateaued.
• Significant differences were observed between lab and home-recorded data.

Conclusions:

• Lumbar-mounted wearables can sensitively detect iRBD-associated motor patterns at home.
• Multi-night home monitoring offers a viable alternative to lab-based vPSG.
• This technology could support staged screening and cohort enrichment for neurodegenerative disease research.