Actimetry in infant sleep research: an approach to facilitate comparability

Area of Science:

• Pediatric Sleep Research
• Biomedical Engineering
• Developmental Psychology

Background:

• Infant sleep behavior exhibits significant variability, necessitating standardized objective assessments for clinical and research purposes.
• Actimetry offers a cost-effective method for objectively measuring infant sleep/wake patterns via limb movements.
• Current actimetry methods face limitations due to varying computational approaches and a focus on nighttime sleep, neglecting significant daytime sleep periods.

Purpose of the Study:

• To evaluate the comparability of two common infant actimetry algorithms (Sadeh and Oakley/Respironics).
• To propose adjustments to enhance the agreement between actimetry algorithms and parent-reported sleep diaries.
• To improve the standardization and accuracy of objective infant sleep assessment.

Main Methods:

• Longitudinal actimetry data collected from 50 infants over 10 days at 3, 6, and 12 months of age.
• Analysis using two algorithms: Sadeh and Oakley/Respironics, comparing minute-by-minute agreement and Kappa statistics.
• Validation against comprehensive 24-hour parent-reported sleep diaries.

Main Results:

• Initial agreement between uncorrected algorithms ranged from 77%-84%.
• A six-step adjustment significantly increased algorithm agreement to 96%-97% and improved agreement with parent diaries.
• Adjustments reduced the discrepancy in Total Sleep Duration estimation from 4.5 hours to 0.2 hours.

Conclusions:

• Proposed adjustments enhance the comparability of infant actimetry data across studies.
• Integrating parent-reported diaries allows for the inclusion of daytime sleep in objective assessments.
• Standardized, comparable infant sleep data supports the development of normative trajectories and clinical cutoffs.