Conference Proceeding


Correspondence between actigraphy and PSG measures of sleep onset latency in young children Public Deposited

Downloadable Content

Download PDF
  • INTRODUCTION: Actigraphy is a non-invasive tool providing objective measurement of sleep onset, offset, and efficiency for extended periods of time based upon wrist-activity levels. Whether actigraphy may also provide an adequately-valid estimate of sleep-onset latency (SOL) in young children is not well-established. This study examined concordance between the gold standard of SOL, polysomnography (PSG), and actigraphy in a cohort of 2-5 year-olds studied at five different levels of prior wakefulness.

    METHODS: Participants were 8 healthy children (3 males) studied at three longitudinal time points (2.5-3.0y, 3.5-4.0y, 5.5-6.0y). Children followed a strict sleep schedule for at least 5 days before each of five home-based, PSG recordings in which they also wore an actigraph (AW64). Sleep assessments occurred after 4h, 7h, 10h, 13h, and 16h of prior wakefulness, reflecting different levels of sleep pressure. Visual stage scoring used 30-sec epochs from C3/A2. Lights-out time was simultaneously marked on PSG and actigraphy with event markers. Sleep-onset was the first epoch of stage 2 sleep (PSG) and the first of three consecutive epochs of scored sleep after lights-out (actigraphy).

    RESULTS: Analysis included 9-14 sleep assessments per child of SOL (concurrent PSG and ACT). Averaged SOL varied across sleep assessments and age (PSG range: 4.9+3.1 to 26.9+13.7; ACT range 4.2+3.1 to 19.3+15.9). We performed a nested correlation between PSG- and actigraphy-derived measures of SOL, covarying sleep pressure and age of assessment, nested within subject. The median partial correlation was r=.874 (p<.001), with a range of r=.243 to r=.969. Two children had very-low, non-significant correlations resulting from an outlier in which actigraphy underestimated SOL.

    CONCLUSION: Overall, these findings suggest actigraphy has adequate validity for estimating SOL in young children when using tightly-controlled data collection and analysis procedures. Future analyses should address methods for establishing the minimum number of nights for a reliable estimate of SOL.

Date Issued
  • 2012-06-12
Conference Name
Conference Location
Academic Affiliation
Last Modified
  • 2022-11-17
Resource Type
Rights Statement
  • Craven HJ, Seifer R, LeBourgeois MK. Correspondence between actigraphy and PSG measures of sleep onset latency in young children. Poster presented at: 26th Annual Meeting of the Associated Professional Sleep Societies; June 12, 2012; Boston, MA.