If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
This study longitudinally compared the sleep of infants in the United States whose mothers were in home confinement to those whose mothers were working as usual throughout the COVID-19 pandemic.
Methods
Mothers of 572 infants (46% girls) aged 1-12 months (M = 5.9, standard deviation = 2.9) participated. Assessments were conducted on 4 occasions from late March to May 2020. Infant sleep was measured objectively using auto-videosomnography. Mothers reported their sheltering status, demographic characteristics, and infant sleep.
Results
Infants of mothers in home confinement had later sleep offset times and longer nighttime sleep durations, compared to infants of mothers who were working as usual. At the end of March, these infants also had earlier bedtimes, more nighttime awakenings, and more parental nighttime visits, but differences were not apparent during April and May.
Conclusions
Living restrictions issued in the United States may have led to longer sleep durations and temporary delays in sleep consolidation for infants of mothers in home confinement.
The COVID-19 outbreak is having a massive impact on human lives. By the end of March 2020, most people in the United States were asked to “shelter in place,” with exceptions limited mostly to essential workers continuing their work outside of the house.
Dealing with sleep problems during home confinement due to the COVID‐19 outbreak: practical recommendations from a task force of the European CBT‐I Academy.
However, no data have yet indicated how COVID-19 home-confinement has affected infant sleep.
Sleep-wake patterns undergo dramatic evolution during the beginning of life. The multiple sleep episodes distributed across the 24-hour day in most newborn infants, gradually consolidate into one major nocturnal sleep episode towards the end of the first year.
Parents determine sleep timing and setting, and provide external regulation to facilitate the transition from wakefulness to sleep. Correspondingly, there is abundant evidence linking parents’ emotions, cognitions, and daily routines with their infants’ sleep patterns.
Disruption in family routines may compromise infant sleep consolidation, as demonstrated by a recent study, linking household chaos with delays in infant sleep consolidation processes.
The major disruption to daily routines and stressful circumstances that COVID-19 stay-at-home orders have introduced in households across the United States may have similarly impacted infant sleep in the midst of this pandemic.
Using objective sleep data, this study aimed to longitudinally compare the sleep of infants in the United States whose mothers were in home confinement to those whose mothers were essential workers, working as usual (EWAU), from late March (when “shelter in place” restrictions were issued), and through the beginning of May 2020. We hypothesized that infants of mothers in home confinement would show more fragmented sleep compared to infants whose mothers were EWAU. Considering the rapid evolution in sleep as infants grow in age, and their high dependency on mothers for sleep regulation early in life,
we expected that greater disruption in sleep associated with home confinement would be found in younger compared to older infants.
Methods
Participants
Mothers of 572 infants (46% girls) aged 1-12 months (Mean = 5.9, Median = 6.0, standard deviation = 2.9) participated in this study. Participants resided across the United States (for a map depicting the distribution of participants across states, see supplementary Fig. S1). Table 1 presents baseline demographic characteristics for the entire sample, and separately for infants of mothers in home confinement and EWAU.
Table 1Baseline demographic characteristics of infants and mothers presented for the entire sample, and separately for mothers in home confinement and essential, working as usual
Users of the Nanit baby monitor (described below) were invited via email to participate in a series of 4 online surveys about infant sleep throughout the COVID-19 pandemic. Survey links were sent once every 2 weeks between March 26 and May 7, 2020, and left active for 4 days. At each assessment, objective sleep data were collected using a computer-vision algorithm from the consecutive 10-day period prior to survey completion. Data were collected anonymously, and informed consent was obtained electronically prior to the first survey. IntegReview institutional review board approved the study (Nanit 2017-01; integreview.com). Participants who completed all 4 surveys were offered a respondent reward (raffle prize for 1 of 5 $100 gift cards).
Measures
Auto-videosomnography data. Infant sleep was assessed objectively using Nanit camera monitors in the naturalistic home-setting. These devices continuously record motion within the crib throughout a predefined nocturnal period. A computer-vision algorithm then translates motion-stillness patterns into sleep-wake patterns. Derived metrics have been previously validated against actigraphy as well as polysomnography.
The following metrics were used for this study: (1) number of nighttime awakenings; (2) number of parental nighttime crib visits; (3) sleep duration (total minutes scored as sleep within the sleep period); (4) sleep onset time, defined as the first minute of 5 consecutive minutes of sleep; and (5) sleep offset time, defined as the first minute of wakefulness after the last sleep segment before the infant is taken out of the crib.
Parent reported data. At baseline, mothers completed a demographic online questionnaire, which included items regarding themselves and their infants (e.g., age, ethnicity). Additionally, at each of the 4 online surveys, they reported their employment status, and whether their infant breastfed or shared their bedroom in the past week. They also reported the average duration of infant daytime sleep. To assess sheltering status throughout the study, mothers were asked to indicate at each of the 4 time points whether they were in quarantine, following guidelines to “shelter in place,” essential workers working as usual, or unrestricted due to the coronavirus. Given that the study began a few days after restrictions were imposed in the United States, only N = 2 (0.2%) of mothers endorsed the latter option, and were removed from analyses. A binary measure was created, combining “quarantine” and “sheltering” categories, as compared to “essential, working as usual.”
Data analysis plan
Of the 727 US mothers who completed the baseline survey (including items regarding sheltering status and infant age), objective sleep data for at least 1 assessment was available for 572 infants aged 1-12 months, who comprised the final sample. Data were excluded if sleep onset times were detected outside the 5 PM-12 AM range (N = 9-13, 1.7%-2.5%), and/or sleep offset times were detected outside the 4 AM-9:30 AM range (N = 3-8, 0.5%-1.4%). Specifically, objective data were available for N = 515, N = 534, N = 541, and N = 549 infants at assessments 1-4, respectively. Parent reported data were available for N = 572, N = 334, N = 297, and N = 275 at assessments 1-4, respectively. Attrition was not associated with sheltering status. Outliers were identified using the interquartile range rule,
and winsorized by replacing the outlier with the closest value not identified as an outlier. No values were winsorized for sleep duration, 4-7 (0.7%-1.4%) values were winsorized for parental visits, and 2-3 (0.4%-0.5%) were winsorized for nighttime awakenings at different assessment points.
longitudinal models with an autoregressive correlation structure were fitted for sleep outcomes, including main effects for time (4 assessments) and sheltering status (home confinement vs. EWAU), as well as their interaction terms. To test age moderation effects, models were additionally computed including infant time-by-sheltering status-by-age interaction terms. Significant interaction effects were interpreted using post hoc marginal means pairwise comparisons, for infants younger and older than the median. GEE accounts for repeated within-subject measurements, and accommodates missing data using the full information maximum likelihood approach. Based on significance in univariate analyses, adjusted models were fitted using infant age, mother's ethnicity, marital status, and household income as time-unvarying co-variates, as well as breastfeeding, room-sharing, and employment status as time-varying covariates.
Results
GEE modeling indicated that as time progressed, infants had longer objective nighttime sleep durations (Wald = 18.01, P < .001), fewer nighttime awakenings (Wald = 101.15, P < .001), fewer nighttime parental visits (Wald = 57.24, P < .001), earlier sleep offset times (Wald = 26.20, P< .001) and shorter reported daytime sleep durations (Wald = 64.75, P < .001). These correspond with normal developmental infant sleep trajectories.
Across assessments, infants of mothers in home confinement had later objective sleep offset times (Mdifference = 9.2 minutes, SE = 4.5, Wald = 4.21, P = .04), and longer nighttime sleep durations (Mdifference = 13.9 minutes, SE = 6.5, Wald = 4.59, P = .03), compared with infants of EWAU mothers. Furthermore, significant time-by-sheltering status interaction effects emerged for the number of nighttime awakenings (Wald = 8.65, P = .03), parental nighttime visits (Wald = 13.82, P = .003) and sleep onset time (Wald = 11.07, P = .01; Fig. 1). These indexed that infants of mothers in home confinement had more nighttime awakenings (Mdifference = 0.5, P = .01), more crib visits (Mdifference = 0.6, P = .002) and earlier bedtimes (Mdifference = 20.2 minutes, P = .003) at the end of March, whereas differences were not apparent in later assessments. Furthermore, infant time-by-sheltering status-by-age interaction effects were found significant for parental nighttime visits (Wald = 9.71, P = .02) and reported daytime sleep duration (Wald = 9.95, P = .02). As shown in Fig. 2A, late-March differences between groups in nighttime crib visits were more pronounced in younger compared to older infants. Similarly, younger infants whose mothers were sheltering in place slept for longer durations throughout the day in March and April, whereas differences between groups were not apparent in older infants (Fig. 2B). Other main and interaction effects were nonsignificant.
Fig. 1Objectively measured number of infant nighttime awakenings (A), number of parental nighttime crib visits (B) and infant sleep onset time (decimal) (C) at each of the four assessment points, by sheltering status. P values and mean differences between infants of mothers in home confinement and infants of EWAU mothers are presented for each significant pairwise comparison. Error bars represent standard errors of the mean.
Fig. 2Objectively measured number of parental nighttime crib visits (A) and parent-reported daytime sleep durations (B) at each of the 4 assessment points, by sheltering status and infant age (younger infants were 1-6 months old, older infants were 7-12 months old). P values and mean differences between infants of mothers in home confinement and infants of EWAU mothers are presented for each significant pairwise comparison. Error bars represent standard errors or the mean.
Using objective measurement of sleep and a longitudinal design, this study found that infants of mothers in home confinement woke up 9 minutes later, and slept for 14 minutes longer per night on average, compared to infants of EWAU mothers. Furthermore, at the end of March, these infants also had earlier bedtimes, more nighttime awakenings, and more parental nighttime visits. Interestingly, differences were no longer apparent during April and early May. Age significantly moderated these effects for nighttime visits, with more pronounced differences between groups in younger compared to older infants. In late March, 1–6-month-old infants of mothers in home confinement had approximately 1 more nightly crib visit compared to infants of EWAU mothers, whereas the equivalent difference was smaller (+0.6) in >6-month-old infants. Similarly, daytime sleep duration was significantly longer in infants of mothers in home confinement compared to infants of EWAU mothers in the first 3 assessments, but this was true only for younger infants.
Taken together, our findings index longer, yet less consolidated sleep in infants of mothers in home confinement following COVID-19 restrictions. Such fragmented sleep-wake patterns are characteristic of young infants, whose sleep is usually distributed across the day and night.
The longer sleep periods, indicated by earlier bedtimes and later rise times along with prolonged daytime sleep durations in younger infants, seem to have been “traded-off” with more nocturnal awakenings and visits, possibly by reducing sleep homeostatic pressure.
Differences between groups may have originated from changes in daytime routines associated with home-confinement, as EWAU mothers may have maintained more regular schedules, which have been shown to promote infant sleep.
thus consolidating infant sleep via increased homeostatic pressure. Differences may have also been driven by parental emotional or stress-related factors, associated with being in lockdown.
Such factors have previously been linked with increased parental involvement with the infant's sleep, more infant sleep disruptions (e.g., more frequent awakenings), and less success in improving infant sleep via intervention.
Of note, most differences in sleep metrics were only apparent at the end of March, closely after restrictions were imposed in the United States. The lack of differences between groups in later assessments may be reflective of improved adjustment to the novel situation of home-confinement after a few weeks of experiencing it. This is in line with a recent longitudinal study reporting reductions in psychological impact at 4 weeks, compared to directly following the COVID-19 outbreak in China.
In addition, despite the dramatic circumstances generated by the pandemic, sleep of infants in this study consolidated across time. These findings may demonstrate the plasticity of good sleep of infants when exposed to a novel situation—a finding previously found in adults.
The development of insomnia or the plasticity of good sleep? A preliminary study of acute changes in sleep and insomnia resulting from an analogue trauma.
Interestingly, younger infants were more largely impacted by maternal sheltering status compared to older infants. Infants aged 1-6 months of mothers sheltering in place slept for 41 minutes longer during the daytime in late March, and had 1 additional nighttime visit on average, compared to infants of EWAU mothers. These pronounced effects may be due to a greater dependency on mothers early on in a child's life,
as well as by a relative lack of structure for younger infants. Structured schedules may have already been established for older infants, making their daytime sleep-wake routines more durable in the face of changes in maternal routines.
Given the importance of sufficient good quality sleep for the well-being of both infants and their parents, our findings hold substantial clinical importance. Longer sleep duration has been associated with a myriad of beneficial outcomes in young children, including improved health, cognitive functioning, and emotion and behavior regulation.
Chaput J-P, Gray CE, Poitras VJ, et al. Systematic review of the relationships between sleep duration and health indicators in the early years (0–4 years). 2017;17(5):855.
While research delineating precise dose-response effects in infants are lacking, the impact of additional sleep may be even stronger in this population, due to the vast developmental processes occurring during sleep in infancy.
Thus, the additional 14 minutes of sleep per night found in infants of mothers sheltering in place, as well as the additional 41 minutes of daytime sleep found in younger infants of home confined mothers, may have had substantial advantageous implications. On the other hand, our findings revealed more fragmented sleep in infants of mothers sheltering in place following COVID-19 restrictions, as well as more parental crib visits. Sleep fragmentation has been linked with adverse outcomes for both infants and parents.
Considering the “trade-off” between longer and more fragmented sleep that extended sleep opportunities may present, balancing sleep quantity and quality may pose a challenge for parents in the face of changing circumstances, such as those presented by COVID-19 restrictions.
Several limitations merit consideration. First, we were not aware of the type of work mothers were performing. Given recent evidence of increased sleep problems in healthcare workers throughout COVID-19,
differentiating between mothers working in the healthcare system (e.g., doctors/nurses), and essential workers from other fields (e.g., postal/grocery store workers) could have yielded a more detailed representation of infant sleep as a function of their mothers’ sheltering/working status. Second, data were not collected regarding the sheltering status of mothers’ domestic partners or regarding the type of infant care while mothers were at work, limiting our understanding of broader environmental influences on infant sleep. For example, differences may have been found between infants whose parents were both confined to the home, compared to those with one or both parents working outside the house. Furthermore, associations between maternal and infant sleep could have shed light on the mechanistic processes underlying the effects of home confinement on sleep of infants and their mothers, and future studies may wish to investigate these links. Finally, the relatively homogeneous sample, representative mostly of middle-upper socioeconomic status in the United States that have access to these camera monitors, limits the generalizability of our findings. Further work is warranted to examine the impact of home confinement during the COVID-19 pandemic on infant sleep in more diverse populations.
Notwithstanding these limitations, to the best of our knowledge, this study is the first to present longitudinal objective findings regarding infant sleep patterns during the COVID-19 pandemic. Findings suggest that living restrictions issued toward the end of March in the United States may have led to longer sleep durations, as well as temporary regressions or delays in sleep consolidation for infants of mothers in home confinement, which subsided as time progressed. As these unprecedented circumstances unfold—or additional “waves” occur—more work is required to understand the influences of lockdown measures on infant sleep, and their underlying mechanisms.
Declaration of conflict of interest
(a) Financial disclosure: MK and NB have served as consultants for Nanit. AG was an employee of Nanit at the time of study implementation. MG has served a Pro-Bono consultant for Nanit.
(b) Nonfinancial disclosure: None.
Funding
This study was supported by Nanit.
Acknowledgments
The authors wish to thank the participating families for their contribution to the study.
Dealing with sleep problems during home confinement due to the COVID‐19 outbreak: practical recommendations from a task force of the European CBT‐I Academy.
The development of insomnia or the plasticity of good sleep? A preliminary study of acute changes in sleep and insomnia resulting from an analogue trauma.
Chaput J-P, Gray CE, Poitras VJ, et al. Systematic review of the relationships between sleep duration and health indicators in the early years (0–4 years). 2017;17(5):855.
P < .05.
30293-X&id=gr1.jpg){kind=link}
30293-X&id=gr2.jpg){kind=link}