Objectives:
To compare different actigraphy scoring settings with polysomnography (PSG) for 1 night of total sleep time (TST), sleep efficiency (SE), wake after sleep onset (WASO), and sleep onset latency (SOL) in healthy pregnant women between 6 and 7 months of gestation.
Design:
Secondary analysis using data from a case-control study.
Setting:
A large university-affiliated hospital in the Midwestern United States.
Participants:
A total of 78 pregnant women were recruited, among which 38 healthy women with uncomplicated pregnancies were included for this analysis.
Measurements:
Participants had an overnight PSG assessment at a sleep center while simultaneously wearing an actigraph (Minimitter; Philips Respironics, Andover, MA). Sleep parameters from both devices included TST, SE, WASO, and SOL. Four scoring settings were used to obtain these parameters from actigraphy. Bland-Altman tests were used to evaluate the agreement between sleep variables scored independently from actigraphy and PSG.
Results:
The default scoring setting (10-by-40) yielded significantly different results from the PSG (P < .01). The 10 immobile/mobile minutes for sleep onset/end with an activity threshold of 10 (10-by-10) produced estimations of TST, SE, and WASO closest to those produced by PSG. When this setting was used, the mean differences between PSG- and actigraphy-assessed TST, SE, and WASO were −1.9 minutes, −0.4%, and 7.4 minutes. When the 10 and 15 immobile/mobile minutes for sleep onset/end were used, the difference between PSG- and actigraphy-assessed SOL was approximately 4 to 5 minutes.
Conclusions
Findings from this study do not support the use of default actigraph settings in pregnant women. In contrast, the 10-by-10 scoring setting provided the greatest agreement and least bias in comparison with PSG for sleep measurements. The 10-by-10 scoring setting is recommended to be used in studies consisting of pregnant women.