Patient population
Of the 105 surveys started, 4 were omitted because no responses were entered past the first question. The majority of the respondents’ children were less than 9 years old (65%) and lived in a home with two parents (86.1%). Almost half were in private school, 35% in public and 12% were home-schooled before the pandemic (Table 1).
Table 1
Description of the children of survey respondents
Variable | Categories | Total population (n=101) |
Mean age (years) | 8.5 (±2.6) |
Male sex n(%) | 52 (52%) |
Home family structure | Single parent | 5 (5%) |
2 parent home | 87 (8%) |
Split time between homes | 7 (7%) |
Income level | $25,000-99,999 | 26 (26%) |
$100-149,999 | 30 (30%) |
$150-249,999 | 22 (22%) |
>$250,000 | 15 (15%) |
Parent is an essential worker | 61 (60%) |
Essential worker is in health care | 26/51 (26% of total population) |
School type before the pandemic | No school yet (< 5 years old) | 5 (5%) |
Public | 35 (35%) |
Private | 47 (47%) |
Home schooled | 12 (12%) |
Data are presented as mean (±SD) or count (proportion of column) unless otherwise specified. |
Total sleep time
Average total sleep time (TST) per night was 10.34 (±0.79) hours before the pandemic, 10.76 (±1.07) hours during lockdown and 10.36 (±0.98) hours when the school year started in fall 2020. Average TST increased by 0.42 hours or 25 minutes (95%CI 0.21-0.64, p<0.001) from before the pandemic to during lockdown. Average TST then fell by 0.36 hours or 22 minutes (95%CI 0.12-0.60, p=0.003) from lockdown to the start of the school year. There was no significant difference between TST before the pandemic and during the school year (difference 0.04 hrs, 95%CI -0.11 to 0.19, p=0.613).
There was a negative linear association between age and TST before the pandemic (r=-0.458, p<0.001). This correlation lessened during lockdown, r=-0.168, p=0.051 and returned when the school year started again (r=-0.377, p<0.001). This trend suggests that older students slept more on lockdown but when school restarted, they woke up sooner, shortening the TST.
Bedtime and wake times
The average bedtime before the pandemic was 20:31 (±0:46) on weekdays and 21:11 (±1:06) on weekends. This increased to 21:01 (±2:22) on lockdown weekdays and 21:46 (±1:36) on lockdown weekends. When school restarted in the fall 2020, the average bedtime was 20:52 (±1:01) on weekdays and 21:26 (±1:18) on weekends.
There was a statistically significant change in the average bedtime to 29 minutes later on weekdays (0:29, 95%CI 0:00-0:58, p=0.046) and 34 minutes later on weekends (0:34, 95%CI 0:23-0:46, p<0.001) when comparing before the pandemic to the lockdown period. When comparing the average bedtime during lockdown to when school started in the fall of 2020, weekday bedtime did not change significantly (-0:07, 95%CI -0:37 to 0:22, p=0.636), but weekend bedtime did occur 18 minutes earlier (-0:18, 95%CI -0:34 to -0:02, p=0.024) when school started. Overall, the average bedtime on weekdays was 21 minutes later when school restarted then compared to before the pandemic (0:21, 95%CI 0:14-0:28, p<0.001) and 15 minutes later on weekends (0:15, 95%CI 0:03-0:27, p=0.010).
The average wake time for all subjects before the pandemic was 6:45am (±0:42) on weekdays and 7:47am (±1:17) on weekends. This increased to 8:01am (±1:32) on weekdays and 8:19am (±1:44) on weekends during lockdown. When school started, the average wake time on weekdays was 7:06am (±1:01) and on weekends was 8:06 (±1:36). All comparisons of wake time from before the pandemic to during the lockdown to during the school year on weekends and weekdays was statistically significant (all p values 0.001 or less).
Parental reports of sleep habits
The majority of respondents reported that before the pandemic their children were doing well with going to bed at the same time every night (87%), sleeping independently (82%), not struggling at bedtime (89%), not having nightmares (92%) and getting adequate sleep (68%). Every aspect of these subjective assessments of sleep habits worsened during lockdown and did not return to pre-pandemic levels during the fall of 2020 (Table 2).
Table 2
Parental report of school-aged children’s sleep quality and behaviors before COVID-19, during the school closures (lockdown) and during the school year in the fall of 2020.
Sleep habit | Before the pandemic | Lockdown | School year in fall 2020 | Comparing before to lockdown | Comparing lockdown to fall 2020 | Comparing fall 2020 to before |
USUALLY or ALWAYS Went to bed the same time every night | 87% | 39% | 58% | 0.025* | <0.001* | <0.001* |
RARELY Slept in another’s bed | 82% | 69% | 72% | <0.001* | <0.001 | <0.001 |
RARELY Had a nightmare | 92% | 67% | 75% | 0.002* | <0.001 | 0.003* |
RARELY Struggled at bedtime | 89% | 65% | 70% | <0.001* | <0.001* | 0.014* |
RARELY Tired or hyperactive during the day | 65% | 40% | 43% | <0.001* | <0.001 | <0.001* |
RARELY Did not get enough sleep | 68% | 46% | 47% | <0.001* | <0.001 | <0.001* |
Had GOOD or EXCELLENT Sleep | 90% | 48% | 58% | 0.060* | <0.001 | 0.016* |
Data are shown as the percentage of respondents who chose the frequency of each behavior described on the left most column. Chi-square testing used unless cell count mandated Fisher’s Exact Testing.* |
Before the pandemic, 9 in 10 parents reported that their child’s sleep was good or excellent. This dropped to 48% of parents during lockdown and increased to 58% during the start of the 2020 school year.
Reported screen time
The proportion of parents who reported their child had more than 4 hours of screen use per day went from 6% before the pandemic to 67% during lockdown and to 40% once school restarted in the fall of 2020. Screen use for more than 8 hours a day occurred in 0% of children before the pandemic, in 19% during lockdown and 16% in the fall of 2020 (Figure 1). The change from before the pandemic to during the school year was statistically significant (p=0.003) as was the decrease from lockdown to when school started (p<0.001). Only 13% of parents reported that screen use before the pandemic was related to education, this increased to 52% during lockdown and 43% during the school year.
Comparing types of schools
The breakdown of demographics and sleep parameters for the children by school type before the pandemic and during lockdown are described in Table 3. Before the pandemic, homeschoolers slept longer (p=0.057) and woke up later when compared to those in private, public or not in school yet (p=0.006). Overall the bedtime and wake times varied less between weekends and weekdays in children who were home-schooled than those in public or private school before the pandemic. A pattern that persisted even during lockdown. The differences by school type in total sleep time and wake time did disappear once the schools were closed. Of note, the average bedtime during lockdown for those children in public school was significantly later (22:20 ±1:44) on weekdays than that of children in private schools (21:44 ±1:31) or home-schooled (20:50 ±0:39), p=0.010.
Table 3
Comparison of children as categorized by school-type before the COVID-19 pandemic began
Variable | Categories | Not in school yet (n=5) | In public school (n=35) | In private school (n=47) | Home schooled (n=12) | p-value |
Mean age (years) | 5.2 (±0.45) | 9.3 (±2.9) | 8.4 (±2.5) | 8 (±1.8) | 0.004 |
Male sex n(%) | 3 (60%) | 20 (57%) | 20 (44%) | 9 (75%) | 0.226 |
Home family structure | Single parent | 0% | 3 (9%) | 2 (4%) | 0% | 0.377 |
2 parent home | 5 (100%) | 31 (89%) | 39 (83%) | 12 (100%) |
Split time between homes | 0% | 1 (3%) | 6 (13%) | 0% |
Income level | $25,000-99,999 | 1 (20%) | 9 (27%) | 13 (28%) | 3 (25%) | 0.446 |
$100-149,999 | 1 (20%) | 11 (32%) | 11 (23%) | 7 (58%) |
$150-249,999 | 2 (40%) | 6 (18%) | 14 (30%) | 0% |
>$250,000 | 1 (20%) | 7 (21%) | 7 (15%) | 0% |
Parent is an essential worker | 2 (40%) | 25 (74%) | 25 (56%) | 9 (75%) | 0.205 |
Essential worker is in health care | 1 (20%) | 12 (35%) | 11 (24%) | 2 (17%) | 0.957 |
SLEEPING PATTERNS BEFORE THE PANDEMIC |
Bedtime on weekdays | 19:54 (±0:25) | 20:42 (±0:48) | 20:29 (±0:49) | 20:30 (±0:28) | 0.157 |
Bedtime on weekends | 20:12 (±0:50) | 21:38 (±1:09) | 21:05 (±1:05) | 20:50 (±0:39) | 0.009 |
Wake time on weekdays | 6:36 (±0:32) | 6:42 (±0:47) | 6:36 (±0:31) | 7:23 (±0:44) | 0.006 |
Wake time on weekends | 7:12 (±0:54) | 8:08 (±1:30) | 7:40 (±1:14) | 7:41 (±0:49) | 0.252 |
Total sleep time on weekdays (hours) | 10.7 (±0.76) | 10.0 (±1.1) | 10.2 (±0.75) | 10.8 (±0.85) | 0.057 |
Total sleep time on weekends (hours) | 10.8 (±1.15) | 10.4 (±0.9) | 10.5 (±1.0) | 10.9 (±0.84) | 0.538 |
SLEEPING PATTERNS DURING LOCKDOWN |
Bedtime on weekdays | 19:48 (±0:40) | 21:42 (±1:21) | 20:41 (±3:07) | 20:56 (±0:35) | 0.166 |
Bedtime on weekends | 20:12 (±1:05) | 22:20 (±1:44) | 21:44 (±1:31) | 21:08 (±0:52) | 0.010 |
Wake time on weekdays | 7:06 (±0:53) | 8:31 (±1:57) | 7:53 (±1:15) | 7:45 (±1:03) | 0.094 |
Wake time on weekends | 7:18 (±1:12) | 8:54 (±2:01) | 8:12 (±1:35) | 7:50 (±0:56) | 0.069 |
Total sleep time on weekdays (hours) | 11.3 (±1.2) | 10.8 (±1.3) | 10.8 (±0.92) | 10.8 (±1.1) | 0.789 |
Total sleep time on weekends (hours) | 11.1 (±1.5) | 10.6 (±1.2) | 10.5 (±1.7) | 10.7 (±1.2) | 0.832 |
SCREEN USE |
Daily screen time BEFORE the pandemic > 4 hours | 0 | 5 (14%) | 1 (2%) | 0 | 0.088 |
Screen use was for education | 0 | 3 (9%) | 5 (11%) | 5 (42%) | 0.017 |
Daily screen time DURING LOCKDOWN > 4 hours | 2 (40%) | 28 (82%) | 35 (75%) | 1 (8%) | <0.001 |
Screen time increased in lockdown | 3 (60%) | 29 (85%) | 45 (96%) | 2 (17%) | <0.001 |
Screen use was for education | 1 (20%) | 22 (65%) | 27 (57%) | 3 (25%) | 0.042 |
Data are presented as mean (SD) or count (column percentage) unless otherwise specified. Comparisons were made using ANOVA for continuous data and Kruskal-wallis for categorical variables. |
While screen time increased in children in all school types, the group with the largest proportion who had increased daily screen time in lockdown were those in private school (96%) when compared to 60% of those not in school yet, 85% of those in public school and 17% of those home-schooled, p<0.001.
School type when the schools re-opened
Of the children in private school before the pandemic, only 7% (n=3) changed school type during the fall of 2020; all went to home schooling. Of the children in public school prior to the pandemic, 26% changed school type in the fall with most (7 out of 9) going to private school where the majority (5 of 7) received all in-person schooling (Figure 2).
Regardless of school type, the majority of students (64%) were in hybrid, virtual or home-schooled when the school year started in fall of 2020. Table 4 shows the comparison between children who were home-schooled, those receiving hybrid or all virtual classes (listed as hybrid) and those receiving all in-person education. The average amount of sleep tended to be higher in home-schoolers when compared to those in hybrid or in-person education (10.9 hours versus 10.3 and 10.2, p=0.057) as wake time was later on weekdays (7:42 ±1:01 vs 7:14 ±1:09 and 6:41 ±0:36, p=0.042) respectively.
Table 4
Sleep characteristics and screen time in children attending in-person, hybrid or being home schooled during the beginning of the 2020-2021 school year.
| Home-schooled (n=17) | Hybrid/virtual (n=45) | All in-person (n=35) | p-value |
Mean age (years) | 8 (±1.9) | 8.8 (±2.8) | 8.4 (±2.8) | 0.595 |
Average daily hours of sleep | 10.9 (±0.8) | 10.3 (±1.1) | 10.2 (±0.8) | 0.057 |
Average hours of sleep on weekdays | 10.9 (±0.8) | 10.2 (±1.4) | 10.1 (±0.9) | 0.241 |
Average hours of sleep on weekdays | 10.9 (±0.9) | 10.7 (±1.1) | 10.5 (±1.1) | 0.794 |
Bedtime on weekdays | 20:49 (±0:35) | 21:06 (±1:13) | 20:36 (±0:50) | 0.680 |
Wake time on weekdays | 7:42 (±1:01) | 7:14 (±1:09) | 6:41 (±0:36) | 0.042 |
Bedtime on weekends | 21:07 (±0:48) | 21:38 (±1:25) | 21:24 (±1:19) | 0.211 |
Wake time on weekends | 7:58 (±1:12) | 8:17 (±1:45) | 7:59 (±1:36) | 0.546 |
USUALLY or ALWAYS Went to bed the same time every night | 10 (59%) | 21 (48%) | 24 (69%) | 0.175 |
RARELY Slept in another’s bed | 16 (94%) | 30 (67%) | 24 (69%) | 0.083 |
RARELY Had a nightmare | 16 (94%) | 34 (76%) | 22 (63%) | 0.052 |
RARELY Struggled at bedtime | 14 (82%) | 28 (62%) | 26 (74%) | 0.241 |
RARELY Tired or hyperactive during the day | 14 (82%) | 17 (39%) | 11 (31%) | 0.002 |
RARELY Did not get enough sleep | 14 (82%) | 15 (34%) | 16 (46%) | 0.003 |
Sleep quality was GOOD or EXCELLENT | 14 (82%) | 20 (44%) | 22 (63%) | 0.020 |
Screen time > 4 hours per day | 2 (12%) | 34 (76%) | 3 (9%) | <0.001 |
Screen time > 8 hours per day | 0 | 14 (31%) | 1 (3%) | <0.001 |
Data are presented as mean (±SD) or count (column percentage). Comparisons between continuous data were made using ANOVA and that between categorical variables using Kruskal-wallis testing. |
A larger proportion of home-schoolers slept well overall. Children in hybrid or in-person schooling had lower rates of getting enough sleep and were more often tired or hyperactive during the day. Per parental report, children in hybrid or virtual learning had the lowest rates of good or excellent sleep quality at 44.4% when compared to that of in-person learners (62.9%) and home-schooled children (82.4%), p=0.020. This equates to home-schoolers having 4-fold higher odds of quality sleep when compared to their peers (OR 4.22, 95%CI 1.13-15.84). This association remained significant even after adjusting for daily screen time over 4 hours and having good sleep quality before the pandemic (aOR 5.64, 95%CI 1.09-29.08).
As expected by the educational model, more children in hybrid or virtual classes had daily screen time that exceeded 4 hours (76%), and even 8 hours per day (31%) when compared to home-schoolers (12%, 0%) and to those in all-person learning (9%, 3%), p<0.001. Children with daily screen time less than 4 hours had double the odds of good sleep quality (OR 2.66, 95%CI 1.15-6.14).