Validity of an algorithm for determining sleep/wake states using FS-760 in school-aged children

Area of Science:

• Biomedical Engineering
• Sleep Science
• Pediatric Health

Background:

• Actigraphy, using accelerometers, estimates sleep/wakefulness but requires device-specific algorithms.
• Existing algorithms for the FS-760 actigraph are validated for adults, not children.
• Establishing a validated algorithm for children is crucial for accurate sleep assessment in this demographic.

Purpose of the Study:

• To develop and validate a sleep/wake discrimination algorithm for the waist-worn FS-760 actigraph in school-aged children.
• To evaluate the performance of the existing adult algorithm when applied to children's activity data.
• To identify factors influencing the agreement rates of the adult algorithm in children.

Main Methods:

• Sixty-five healthy school-aged children underwent simultaneous polysomnography and FS-760 actigraphy.
• A five-order linear discriminant analysis was used to formulate the children's sleep/wake algorithm based on activity intensity.
• Adult and children's data were combined for multiple regression analysis to examine algorithm agreement factors.

Main Results:

• The developed algorithm achieved a mean agreement rate of 91.0% for sleep/wake detection in children.
• The adult algorithm showed significantly lower agreement (81.8%) when applied to children's data.
• Factors like mean activity during NREM/REM sleep, age, and sleep efficiency influenced the adult algorithm's agreement rates in children.

Conclusions:

• The newly developed algorithm reliably assesses sleep/wake states in school-aged children using the FS-760 actigraph.
• Adult algorithms demonstrate reduced accuracy in children due to differing activity patterns.
• Population-specific sleep/wake algorithms are necessary for accurate actigraphy-based assessments in pediatric populations.