The Influence of Parental Sleep and Experiences Related to COVID-19 on Sleep in Children and Adolescents between 2020 and 2021 in Brazil

The study aimed to evaluate the influence of parental sleep and experiences related to COVID-19 on sleep changes in children and adolescents in the period between 2020 and 2021 in Brazil and further compare the sleep of adults with and without children. This is a longitudinal web-survey study. Participants were invited to respond to a questionnaire regarding sleep characteristics, mental health issues, and work/lifestyle modifications in two waves of the pandemic (April–July 2020 and 2021). A total of 1172 adults answered both questionnaires, and 281 were dyads (parent–child/adolescent). Parent and non-parent adult responders had similar sociodemographic data, with a predominance of the female sex and self-declared white skin color prevailing along with higher levels of education in both groups. The prevalence of sleep problems in adults varied from 20.6% to 30.2% in the parent group and from 16.9% to 30.1% in non-parents. The prevalence of sleep problems in children and adolescents raised from 2020 to 2021 (respectively, 48% and 49.5%) but differences were not statistically significant. The multivariate logistic model showed in both years that changes in children’s/adolescents’ sleep was related to parents working at home, infected family/friends, time of exposure to COVID-19, and daytime sleep dysfunction. Our data showed that parental sleep and experiences related to COVID-19 influenced sleep changes in children and adolescents. Parents had a significant difference in daytime sleepiness compared to the group without children.


Introduction
At the end of December 2019, new cases of a severe acute pneumonia of unknown origin appeared in Wuhan City, Hubei Province, China. The newly discovered virus belonging to the Coronaviridae family became known as SARS-CoV-2 and quickly spread across all continents [1]. In March 2020, the World Health Organization (WHO) declared a state of public calamity for the pandemic. Characterized by its high transmissibility, the disease caused by the virus was called COVID-19 [1]. More than two years later, it is estimated that the worldwide mortality associated with the COVID-19 pandemic was 14.9 million between 2020 and 2021 [2].
In order to contain the rapid spread of SARS-CoV-2, measures such as the use of protective masks and social isolation started worldwide. In Brazil, social isolation due to COVID-19 started in the second week of March 2020. Only essential services remained open. The abrupt closure caused schools to have their operations interrupted and adapted to the online model [3].

Materials and Methods
This is a longitudinal nationwide web-survey study conducted by the COVID-19 Task Force of the Brain Institute (BraIns) of the Pontifical Catholic University of Rio Grande do Sul, Brazil. Data were collected using the Qualtrics ® online survey software platform (www.qualtrics.com). Participants were invited through the "virtual snowball" sampling method which began by sending invitations through social networks and internet, radio, and TV in two periods, the first between 27 April and 30 July 2020, comprising the first wave of the COVID-19 pandemic in Brazil, and the second period from 5 April to 30 July 2021. The model of social isolation imposed by COVID-19 varied during data collection, where in 2020, Brazil was in complete lockdown, and in 2021, social isolation was partial.
After a brief explanation regarding the aim of the study, participants who agreed to join clicked on the option "I agree to participate in this study" for online informed consent and were then directed to a sociodemographic questionnaire and questions related to COVID-19. Afterward, they were directed to answer questionnaires related to sleep, and participants who had children under 18 years old answered sleep questionnaires regarding their children. All the responses were anonymous and without any other identification of the participants. However, we had asked to leave an e-mail contact if the responder was interested in being surveyed in the second phase of the study. The information was directly stored in the Qualtrics platform and was available only to the senior researcher under the use of a password.
The national survey had a total of 5007 respondents in 2020 and 1582 in 2021, of these, 1172 participated in both years and constitute the sample analyzed in this work. The final sample was divided into adults with children (parents) and non-parents. The present study was approved by the Institutional Ethics Committee and registered on Plataforma Brasil under number 30748320.5.0000.5336.

Instruments
The evaluation of sleep during the survey was based on previously validated questionnaires. The Brazilian version of the "Pittsburgh Sleep Quality Index (PSQI)" was applied to parents and adolescents (13 to 17 years and 11 months). It has seven components that together lead to a global score. In the global score, a score equal to or greater than 5 distinguishes between poor and good sleep quality, and above 10 points indicates a predisposition to sleep disorders [8]. The Brazilian version of the "Epworth Sleepiness Scale (ESS)" was used only in adults. The ESS is an eight-item survey where participants score, on a Likert scale of 0-3, the probability of napping or falling asleep in life activities daily. Out of a total of 24 points, a score > 10 indicates mild or excessive daytime sleepiness [9].
The "Brief Infant Sleep Questionnaire (BISQ)" was used for children aged 0 to 3 years and 11 months. For the BISQ, the cutoff score was one (or more) of the following: more than three nocturnal awakenings, nighttime wakefulness greater than one hour, and/or total sleep time less than nine hours [10].

Comparison between Parents and Non-Parents
The PSQI and ESS scales were compared between the groups (parents x non-parents) in each year of investigation. We sought to identify whether the observed proportions of sleep problems and each of the components of the scales differed between the groups, as well as in each year of the evaluation.

Statistical Analysis
Data were analyzed using the Statistical Package for Social Sciences version 20.0 (SPSS Inc., Chicago, IL, USA, 2008) for Windows, with a significance level of 5% for statistical decision criteria. Results were presented using descriptive statistics-absolute and relative distribution (n-%), as well as mean, standard deviation, and amplitude. The study of age symmetry was carried out using the Shapiro-Wilk test. The analysis that compared the paired proportions between the two years of investigation was the McNemar-Bowker test. Since, in the evaluation of the impact observed between the proportions of each year, the size of the effect was calculated through Cramer's V, considering the classification criterion: <0.10 = insignificant effect; 0.10-0.39 small effect; 0.50-0.69 = average effect; and >0.70 = large effect [12]. In the comparisons between independent groups (gender, age group, and mental health), Pearson's chi-square test (Monte Carlo Simulation) was used.

Logistic Regression
The analysis that involved the prediction of sleep alteration in children and adolescents was investigated using Binary Multiple Logistic Regression. In the selection of independent variables with predictive potential to characterize sleep disorders in children, the Backward conditional method was used and, to verify the goodness of fit of the final logistic regression model, the Nagelkerk and Hosmer-Lemeshow R 2 estimators were considered. The probability of the gradual entry of the variables in the model was used as 0.05 and for the removal of 0.10. Thus, the independent variables in the model were analyzed by the characteristics of the parents who presented a minimum level of significance less than or equal to (p ≤ 0.250) in relation to the child's sleep disorder. In addition, the sex and age range of the children were considered in the model.

Sociodemographic Data
The results presented refer to a sample of 281 parents and 891 adult non-parents investigated in 2020 and 2021, with parental ages ranging from 19 to 51 years (37.3 ± 11.8) and 18 to 91 (38.6 ± 13.9), respectively. The female sex and self-declared white skin color prevailed in the two groups with higher levels of an academic degree. In relation to monthly income, the highest concentrations occurred in Classes A and C. Considering sociodemographic data, no significant differences were observed between the groups. Our sample included responders from all Brazilian regions; however, the largest concentration was from the southern region of the country (76.1%) ( Table 1).

Labor Characterization and Relationship with COVID-19
In Table 2, we described labor characterization and different aspects related to COVID-19; again, no significant differences were observed between the groups. Many responders were health professionals, followed by teachers. Regarding the parents infected, most of them had mild to moderate symptoms (25.3%). Considering issues related to mental health, more than half of the sample reported feeling alone, mentioned that their lives had no meaning, that they self-harmed themselves for some reason, or even had suicidal ideation during the pandemic period.

Parents' and Non-Parents' Sleep
The PSQI and ESS scales were compared between the groups in each year of investigation. Significant differences were observed in PSQI items, such as sleep quality and efficiency. In the subjective component of sleep quality, in 2020, parents claimed worse sleep quality, and in 2021, poor quality was higher in the non-parent group. Regarding sleep efficiency, which was considered normal if >85%, in 2020, parents reported higher efficiency than non-parents. In 2021, there was an inversion in this result, with a decrease in sleep efficiency for those who had children (Table 3).
Regarding the results referring to the ESS scale, significant differences were observed in the items "watching TV", "sitting quietly in a public place", and "Sitting quietly after a lunch without alcohol" (Table 4). In 2020, it was found that parents were more likely to fall asleep watching TV than in 2021 and for those who did not have children. In 2021, parents were more likely to sleep when sitting and quiet in a public place (p < 0.001). Regarding the total score of ESS, in both groups there was an increase from 2020 to 2021, meaning that diurnal somnolence increased in the second year of the pandemic (p < 0.001). As a result, it was noted that in 2020, adults napped less, regardless of whether they were parents or not. However, in 2021 parents had a lower proportion of naps when compared to non-parents.

Sleep of Children and Adolescents
Among children/adolescents included, there was a predominance of the male sex (54.1%) and the majority in the age range between 4 and 12 years (47.3%). In all age ranges, the prevalence of sleep problems was high according to the instruments used, in both years, but without significant differences from 2020 to 2021. However, when parents had sleep disturbances or when children, adolescents, and parents both had sleep disturbances, it proved to be an important aspect in the quality of sleep of children and adolescents (Table 5). OBS: BISQ = Brief infant sleep questionnaire, SDSC = Sleep disturbance scale for children, PSQI = Pittsburg sleep quality index; * estimated by BISQ (0-3 years and 11 months), SDSC (4-12 years and 11 months) and PSQI (13-17 years and 11 months) scales. ** Adults estimated by the PSQI scale (>5 points). *** statistical significance. Table 6 shows the factors associated with the presence of sleep disorders in children and adolescents in 2020 and 2021 using the multivariate logistic model. In 2020, potential predictors of sleep disturbances in children and adolescents were parents working fully from home, parents having been infected by COVID-19, and parents' daytime sleep dysfunction. Adolescents had a higher risk of disorders of sleep compared to the age group of 0 to 3 years. In 2021, predictors of sleep disorders in children and adolescents were related to parents who needed to be exposed to COVID-19 for more than 12 h a day and any family member/close friend who had been infected; parents with sleep efficiency below 85% were more likely to have children with sleep disorders. Children aged 4 to 12 years were more likely to have sleep disorders when compared to those aged 0 to 3 years.

Discussion
In this longitudinal study, it was possible to show that sleep problems persisted in all age ranges across the pandemic. In 2020, both parents and adults who did not have children showed changes in sleep, mainly daytime sleepiness that few said they previously had, but in 2021, daytime sleepiness had a considerable increase. In 2020, adolescents had the highest percentage of sleep disorders, and in 2021, it was school-age children. Regarding aspects related to COVID-19, contagion from close relatives and the exposure time of adults influenced the sleep of children and adolescents.
We know that during the social isolation caused by the COVID-19 pandemic, lifestyle, work, and habits/routines were suddenly changed [13]. Parents found themselves on a mission to work at home and adapt the family to the new routine. In addition, anxiety, fear of the unknown, and economic recession interacted to increase sleep complaints [14]. Even at the beginning of the pandemic, studies showed that parents had difficulty sleeping due to their routines that were also abruptly changed and new family arrangements because of the lockdown [15,16].
It was observed in our study that children and adolescents whose parents worked full-time at home were more predisposed to sleep disorders. This finding corroborates an Italian study carried out with mothers who were working at home, which observed a worsening in the quality of sleep and distortion of the experience of time in mothers and children, in addition to an increase in emotional symptoms and difficulties in self-regulation in children [16].
Studies involving adolescents have shown that sleep disturbances during the COVID-19 pandemic have been associated with increased insomnia, anxiety, and depression, in addition to increased rates of psychiatric symptoms. Loneliness due to social isolation can put children and adolescents at risk for the appearance of psychiatric disorders [17][18][19]. Our study did not explore the emotional aspects of children and adolescents, but adults were asked about aspects of mental health and it was found that more than half of the sample had moments when they felt alone during the first year of the pandemic. Corroborating our study, a multicenter study carried out in 87 countries showed that during the COVID-19 pandemic, negative emotions increased worldwide, potentiating mental illness [20].
As for the contamination of parents by COVID-19 and family members or friends, it influenced the worsening of sleep in children and adolescents. These issues may be linked to feelings experienced, such as the fear of losing a loved one. A study carried out in Canada showed that in 2020, changes in family sleep habits were associated with fears and concerns about COVID-19; in addition, reduced access to health services was also associated with parental concerns [21]. It is important to note that in our study, most respondents were health professionals, so the possibility of exposure to the virus was greater, represented by 40.2% of parents and 22.8% of non-parents. Many studies have associated sleep with the emotional aspects of health professionals, but the literature is scarce when it is focused on the children of these professionals [20,22,23].
As for the limitations of the study, the asymmetric regional distribution of participants with most of the sample from southern Brazil should be mentioned. However, in this region, lockdown protocols were more homogeneous than in other regions of the country. Another limiting aspect is the higher prevalence of respondents with an academic degree, which may make the generalization of the findings to the whole population difficult. As it is known, the pandemic has had different impacts on people from different socioeconomic levels. In addition, the questionnaire allowed parents to respond in relation to only one child, losing control or guaranteeing which criteria the parents used to choose the child they were responding to (the worst sleeper or the youngest, etc.).

Conclusions
Our data showed that during the COVID-19 pandemic, parental sleep and lifestyle issues influenced sleep alterations in children and adolescents across the years 2020 and 2021. Furthermore, it is important to highlight the persistently high rates of sleep disorders in all age groups over the two years, especially in school-age children and adolescents. When comparing sleep characteristics between the two adult study groups, parents had a worse sleep quality.
Author Contributions: The authors S.A.R., C.F.-R., L.B.S. and N.F.C. were responsible for the preparation of the manuscript, collection, analysis, data interpretation, and writing. The author M.L.N. was responsible for the analysis, data interpretation, and final writing. All authors have read and agreed to the published version of the manuscript.