A longitudinal study of sleep-wake patterns during early infancy using proposed scoring guidelines for actigraphy

Highlights

• Change to infants' usual daytime nap sleep duration influenced that nights' sleep.

• Change to infants' usual nighttime sleep interval did not influence next-day sleep.

• Mothers tended to underestimate infants' nighttime sleep compared to actigraphy.

• Mothers tended to overestimate infants' daytime sleep compared to actigraphy.

Abstract

Objective/Background

We describe developmental and day-to-night sleep patterns across the first six months of life using actigraphy and compare these to mother-reported perceptions of infant sleep.

Patients/Methods

This observational, burst design included three, one-week bursts of data collection at six, 15, and 24 weeks of age. Infants wore an actigraphy device (Actiwatch Spectrum) on their right ankle for each one-week period. Data were scored using a SAS-based hierarchical, algorithmic methodology and independently assessed for necessary corrections by two trained scorers in a Visual Basic. Mothers completed the Brief Infant Sleep Questionnaire (BISQ) at each burst. Mixed models tested developmental patterns over time and multilevel models examined day-to-night sleep patterns at each burst.

Results

Daytime nap sleep duration decreased over time (p = 0.02) with marginal significance for nighttime sleep interval duration increasing over time (p = 0.09). Total 24-h sleep duration was time invariant (p > 0.05). These longitudinal patterns were similar when examining mothers' perception of infant sleep. Daily variations demonstrated nighttime sleep interval and maintenance efficiency did not predict next-day nap sleep duration. Yet, at 24 weeks of age, daytime nap sleep was associated with that nights' sleep interval. For every 1-h above infants’ average total daytime nap sleep duration, infants slept ∼15 min longer and 1.0% less efficiently that night (p ≤ 0.05). Mothers overestimated daytime nap sleep and total 24-h sleep, when compared to actigraphy (p < 0.01).

Conclusion

Changes to infants' usual daytime sleep duration impacted subsequent sleep bouts and mothers tended to overestimate infants’ sleep. These patterns should be explored in relation to parenting practices.

Introduction

Infant sleep patterns develop rapidly throughout the first year of life. Shortly after birth, sleep is fragmented and less predictable, as the central circadian pacemaker takes several months to establish the connection from the retina and for light signals to synchronize to the external light and dark cycle [1]. As infants age, sleep consolidation emerges with a more prolonged sleep at night and more sustained wakefulness during the day. Healthy sleep patterns have important implications on later health outcomes including physical growth and the development of a healthy body mass index (BMI) [2], [3]. Therefore, accurate measurements of these infant sleep-wake patterns are essential for evaluating intervention outcomes, providing guidance in clinical care, and understanding how sleep impacts other behaviors such as feeding.

When quantifying infant sleep patterns, much of the existing literature has relied on parent-reported measures, such as questionnaires or sleep diaries [4], [5]. These are biased by subjective report and limited to what parents are able to recall. Instead, actigraphy is a measure in which motor activity is continuously quantified over time using a wrist-worn wearable device. This noninvasive method uses activity counts to infer states of sleep and wake. Actigraphy may be more accurate than parent-report given its objective nature, while also more feasible than polysomnography (PSG) for natural settings like the home environment. Actigraphy has been used widely in sleep research among adults and older children, yet less in infants [6]. Of the studies that include infants, actigraphy has been shown to be a valid measure with agreement rates ranging 72–95% when compared to PSG or observer-identified states of sleep and wake [7], [8], [9].

Like all forms of measurement, actigraphy is not without limitations. The wide variety of devices, scoring algorithms, device placement sites, and definitions of key variables pose a challenge to interpreting results across studies [10]. In particular, there is little guidance available for researchers and clinicians on how to score and interpret infant actigraphy data. This lack of uniformity and clear scoring rules are discussed in a review by Meltzer et al. [6], and Galland et al. [10], with a call for research that clearly describes actigraphy scoring rules and definitions. Therefore, in an effort to advance the evidence base for scoring actigraphy data and improve the rigor and reproducibility of research using these devices for infants, we describe in detail our data scoring rules in our methods.

With the increasing use of actigraphy in pediatric populations [6], it is important to know how sleep values differ when quantified using actigraphy versus mother-reported measures. A few studies have found that mothers report infant sleep schedules (eg, sleep onset/offset) fairly accurately when compared to actigraphy; yet, mothers' perception of infant nighttime sleep duration is less accurate, due to their underestimation of nighttime wakings [11], [12], [13], [14]. One limitation of these studies is that they focus on infants' nighttime sleep behaviors, with little-to-no research that has included infants' daytime sleep behaviors. To explore these relationships, more research is needed that compares actigraphy vs. mother-reported values for both daytime and nighttime sleep durations throughout infancy. Furthermore, the day-to-night and night-to-day temporal patterns of infant sleep have not been studied. On nights when infants sleep longer or more efficient than usual, their next-day nap sleep duration may be affected, and on days when infants' nap sleep duration is longer than usual, that nights’ sleep onset, interval duration, and/or maintenance efficiency (ie, quality) may be affected. These empirical questions remain unanswered and warrant further investigation.

The primary aim of this study is to describe infants' sleep-wake patterns across the first six months of life. To do so, we describe the developmental patterns from 6 to 24 weeks of age and the daily temporal patterns between subsequent night-to-day and day-to-night sleep onset, interval duration, and maintenance efficiency using actigraphy. As a secondary aim, we compare actigraphy and mother-reported values for infants’ daytime nap and nighttime sleep patterns and describe our infant-specific actigraphy scoring process in the methods.