Frontiers in Psychiatry (Feb 2015)

A comparison of actigraphy and sleep diaries for infants’ sleep behavior

  • Wendy A Hall,
  • Sarah eLiva,
  • Melissa eMoynihan,
  • Roy eSaunders

DOI
https://doi.org/10.3389/fpsyt.2015.00019
Journal volume & issue
Vol. 6

Abstract

Read online

Detecting the effectiveness of behavioral interventions to reduce infant night waking requires valid sleep measures. Although viewed as an objective measure, actigraphy has overestimated night waking. Sleep diaries are criticized for only documenting night waking with infant crying. To support potential outcome measure validity, we examined differences between sleep diaries and actigraphy in detecting night waking and sleep duration. We recruited 5.5 to 8-month-old infants for a behavioral sleep intervention trial conducted from 2009-2011. Intervention (sleep education and support) and control groups (safety education and support) collected infant diary and actigraphy data for 5 days. We compared night-time sleep actigraphy with diary data at baseline (194 cases), and 6 weeks (166 cases) and 24 weeks post-education (118 cases). We hypothesized numbers of wakes and wakes of ≥ 20 minutes would be higher and longest sleep time and total sleep time shorter by actigraphy compared with diaries. Using paired t-tests, there were significantly more actigraphy night wakes than diary wakes at baseline (t = 29.14, df = 193, p < .001), 6 weeks (t = 23.99, df = 165, p < .001), and 24 weeks (t = 22.01, df = 117, p < .001); and significantly more night wakes of ≥ 20 minutes by actigraphy than diary at baseline (t = 5.03, df = 183, p < .001), and 24 weeks (t = 2.19, df = 107, p < .05), but not 6 weeks (t = 1.37, df = 156, n.s.). Longest sleep duration was significantly higher by diary than actigraphy at baseline (t = 14.71, df = 186, p < .001), 6 weeks (t = 7.94, df = 158, p < .001), and 24 weeks (t = 17.18, df = 114, p < .001). Night sleep duration was significantly higher by diary than actigraphy at baseline (t = 9.46, df = 185, p < .001), 6 weeks (t = 13.34, df = 158, p < .001), and 24 weeks (t = 13.48, df = 114, p < .001). Discrepancies in actigraphy and diary data may indicate accurate actigraphy recording of movement but not sleep given active infant sleep and self-soothing.

Keywords