Infant sleep is a challenge among parents and a problem seen frequently in pediatric practice. Actigraphy, an activity-based sleep monitoring system, uses a wristwatch-like device with an accelerometer sensitive to body movements. Activity counts derived from actigraphy serve as the basis for sleep-wake determination. Although it has been concluded that actigraphy is a valid measure of sleep during infancy,1,2 external motion, a common experience for infants (eg, being carried or put in an infant swing), is usually overlooked in studies that use actigraphy to assess infant sleep. In some studies, actigraphy records containing external motion were removed prior to analysis2,3; in others, data involving external motion were included,4,5 with most studies not specifying how such data were analyzed. The objective of this study was to assess the activity count and sleep-wake identification accuracy by actigraphy when an infant doll was exposed to external motion.
Methods
Actiwatch-Score (Mini Mitter Respironics, Inc, Bend, Oregon) actigraphs were placed on both ankles of a 20-lb (9.07-kg) infant doll. Twenty-three experiments involving 13 different types of activities (Table) that infants typically experience were conducted by trained investigators (mainly S.-Y. T.). Actigraphic activity counts were stored in 15-second epochs. Repeated trials were performed for the same activity to obtain more reliable results. At the end of each trial, the activity data were scored as asleep or awake at low (wake threshold value=80 activity counts), medium (wake threshold value=40 activity counts), and high (wake threshold value=20 activity counts) sensitivity settings using the Actiware-Sleep 3.4 analysis software (Mini Mitter Respironics, Inc). Data from different trials of the same activity were combined for analysis. The percentages of epochs scored as awake were calculated for the different types of activities. Statistical analysis was performed using SPSS version 14.0 statistical software for Windows (SPSS Inc, Chicago, Illinois). Because the current study involved no human or animal subjects, institutional review board approval was not required from our institution.
Table.
Effect of External Motion on Mean Activity Count and Waking Percentage Scored at Various Sensitivity Thresholds in Recordings Obtained From an Infant Doll
| Activity | Epochs Recorded, No. |
Activity Count, Mean (SD) |
Waking % at Sensitivity Settinga |
||
|---|---|---|---|---|---|
| Low | Medium | High | |||
| Moving automobile, speed 30 mph | 120 | 71.40 (44.33) | 90.83 | 97.92 | 100 |
| Moving automobile, speed 60 mph | 120 | 91.37 (32.25) | 100 | 100 | 100 |
| Moving stroller | 120 | 301.06 (140.70) | 100 | 100 | 100 |
| Moving shopping cart | 60 | 107.45 (105.01) | 100 | 100 | 100 |
| Vibrating in bouncer | 120 | 70.24 (26.33) | 100 | 100 | 100 |
| Infant swing, low speed | 60 | 0 (0) | 0 | 0 | 0 |
| Infant swing, high speed | 60 | 137.80 (50.70) | 92.50 | 95.00 | 96.67 |
| Rocking | 60 | 7.68 (30.95) | 6.67 | 24.17 | 36.67 |
| Bouncing | 80 | 136.99 (96.10) | 100 | 100 | 100 |
| Carrying, ie, BabyBjörn carrier, sling | 180 | 83.50 (70.58) | 86.39 | 93.61 | 97.22 |
| Holding | 100 | 8.50 (40.90) | 13.00 | 18.50 | 29.50 |
| Swaying | 80 | 28.32 (48.84) | 37.50 | 45.63 | 50.63 |
| Burping | 40 | 27.65 (48.96) | 60.00 | 76.25 | 83.75 |
Wake threshold values were 80 activity counts per epoch for the low sensitivity setting, 40 activity counts per epoch for the medium sensitivity setting, and 20 activity counts per epoch for the high sensitivity setting.
Results
The 2 Actiwatches produced similar results as evidenced by the high correlation coefficient (r=0.88) and by the Bland-Altmanplot.6 Therefore, their measurements were combined to provide an overall mean and standard deviation. As seen in the Table, the low-speed infant swing generated an activity count of 0. However, external motion generated by a moving car (at a speed of 60 mph), stroller, shopping cart, and vibrating bouncer as well as bouncing by a caregiver resulted in 100% false awake period identification, even when the actigraphic activity data were scored using low sensitivity settings. Caregivers’ carrying and a high-speed infant swing also produced considerable movement artifacts. Relatively fewer motion artifacts were generated by caregivers’ rocking, holding, swaying, and burping.
Comment
Our results indicate that the external motion typically experienced by infants increases the activity count and influences the accuracy of actigraphy for sleep-wake estimation. The extent to which the external motion influences the sleep-wake scoring is dependent on the type of external motion being performed and where the threshold activity count falls. Results suggest that actigraphy is a reliable and valid method for monitoring infant activity and assessing sleep only when the external motion is carefully and accurately documented. Preliminary results from our current ongoing study conducted in the natural home environment indicate that infants experience different types of external motion for extensive periods of their 24-hour day. For infants who have established a solid nocturnal sleep pattern, actigraphy may be useful given that less parental holding and carrying would occur. Our findings challenge the use of actigraphy for long-term monitoring in infants, which would require precise parental adherence to diary documentation of external motion.
Acknowledgments
Funding/Support: This work was supported by grant P30 NR04001 from the National Institute of Nursing Research.
Footnotes
Author Contributions: Study concept and design: Tsai and Thomas. Acquisition of data: Tsai and Thomas. Analysis and interpretation of data: Tsai and Thomas. Drafting of the manuscript: Tsai. Critical revision of the manuscript for important intellectual content: Tsai and Thomas. Statistical analysis: Tsai and Thomas. Obtained funding: Thomas. Administrative, technical, and material support: Tsai. Study supervision: Thomas.
Financial Disclosure: None reported.
Additional Contributions: Gail Kieckhefer, PhD, assisted with the actigraphy equipment.
Contributor Information
Shao-Yu Tsai, Department of Nursing, College of Medicine, National Taiwan University, Taipei, Taiwan.
Karen A. Thomas, Department of Family and Child Nursing, School of Nursing, University of Washington, Seattle.
References
- 1.Gnidovec B, Neubauer D, Zidar J. Actigraphic assessment of sleep-wake rhythm during the first 6 months of life. Clin Neurophysiol. 2002;113(11):1815–1821. doi: 10.1016/s1388-2457(02)00287-0. [DOI] [PubMed] [Google Scholar]
- 2.So K, Buckley P, Adamson TM, Horne RS. Actigraphy correctly predicts sleep behavior in infants who are younger than six months, when compared with polysomnography. Pediatr Res. 2005;58(4):761–765. doi: 10.1203/01.PDR.0000180568.97221.56. [DOI] [PubMed] [Google Scholar]
- 3.Acebo C, Sadeh A, Seifer R, Tzischinsky O, Hafer A, Carskadon MA. Sleep/wake patterns derived from activity monitoring and maternal report for healthy 1- to 5-year-old children. Sleep. 2005;28(12):1568–1577. doi: 10.1093/sleep/28.12.1568. [DOI] [PubMed] [Google Scholar]
- 4.So K, Adamson TM, Horne RS. The use of actigraphy for assessment of the development of sleep/wake patterns in infants during the first 12 months of life. J Sleep Res. 2007;16(2):181–187. doi: 10.1111/j.1365-2869.2007.00582.x. [DOI] [PubMed] [Google Scholar]
- 5.Jenni OG, Deboer T, Achermann P. Development of the 24-h rest-activity pattern in human infants. Infant Behav Dev. 2006;29(2):143–152. doi: 10.1016/j.infbeh.2005.11.001. [DOI] [PubMed] [Google Scholar]
- 6.Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1(8476):307–310. [PubMed] [Google Scholar]