DOI: https://doi.org/10.21203/rs.3.rs-1700834/v1
Children’s screen time use has increased significantly during the pandemic. Extended school closures and heightened parent stress are associated with children’s behavioural difficulties and time spent watching screens. This cross-sectional longitudinal survey study examined the association between screen time and internalizing and externalizing behaviours in school-aged children (6–12 years) at two time points over the 2020–2021 academic school year. Parents (n = 108) completed survey measures on their parenting styles, stress levels, along with their child’s patterns of screen time and emotional and behavioural difficulties. Children’s average daily screen time was 4.42 hours (SE = 19.04) at baseline and 3.93 hours (SE = 16.35) at 1-year follow up, with no significant change across the school year (p = .12). Increased screen time was associated with a greater incidence of internalizing behaviours in children (p = .03). Children who spent more time on screens and who were in households with parents reporting higher stress levels had increased internalizing behaviours (p = .001). Parent stress was positively associated with children’s externalizing behaviours (p < .001). Children’s screen time use has remained high during the pandemic and is associated with internalizing behaviours. Targeted family intervention plans focused on reducing parent stress and screen time use may aid in improving children’s mental health during the ongoing pandemic.
The COVID-19 pandemic generated unprecedented increases in children and adolescents’ time spent on devices, with some children spending over 6 hours per day on screens (Eyimaya & Irmak, 2021; Nagata et al., 2022; Seguin et al., 2021; Xiang et al., 2020). A parallel epidemic emerged in parents with school-aged children, who reported stress levels in the moderate to high range (Calvano et al., 2021; Li & Zhou, 2021; Seguin et al., 2021). While screen time and parent stress are known to have separate adverse implications on childhood development, little is known about the accumulated risks of prolonged screen time exposure on incidences of internalizing and externalizing behaviours in children, and the moderating influence of parent stress and parenting strategies on these relationships.
Screen time is the amount of time spent using a device with a digitized interface, such as computers, television, video games, and smartphones (Tang et al., 2018). Health care professionals caution that excessive screen time may adversely impact childhood development, health, and mental health outcomes with screen time guideline recommending a maximum of 2 hours of screen time per day for children over the age of five (Canadian Paediatric Society, Digital Health Task Force, 2019; World Health Organization, 2020).
The unparalleled impact of COVID-19 engendered nationwide increases in children’s time spent on devices (Eyimaya & Irmak, 2021; Orgilés et al., 2020; Velde et al., 2021; Xiang et al., 2020). Based on early cross-sectional data in Ontario, children’s average screen time increased from 2.6 hours before the pandemic, to 5.9 hours a day during the initial school closures of the pandemic (Seguin et al., 2021). This daily average is more than double the recommended guidelines for Canadian children. Similar increases occurred in the United States, Turkey, China, Italy, and Spain (Eyimaya & Irmak, 2021; Nagata et al., 2022; Xiang et al., 2020), with concurrent decreases in physical activity also found in some of these samples (Orgilés et al., 2020; Xiang et al., 2020). Factors such as socioeconomic status, race/ethnicity, school satisfaction (for both children and parents), and having a child with special needs, contributed to unregulated screen time in children (Assari, 2020; Lee et al., 2022; Nagata et al., 2022; Seguin et al., 2021).
The most prevalent mental health concerns affecting school-aged children prior to the pandemic were anxiety, depression, and conduct problems (Ghandour et al., 2019). Since March 2020, internalizing (i.e., anxiety and depression) and externalizing problems (i.e., aggression) have increased dramatically in school aged children (Khoury et al., 2021), with an estimated 20–25% of children experiencing symptoms of depression and anxiety (Racine et al., 2021). Lockdowns were associated with the most pervasive emotional problems, including anxiety, restlessness, worry, and depression (Orgilés et al., 2020; Racine et al., 2021).
Early evidence indicates that internalizing and externalizing problems may be an enduring issue for children and youth during the pandemic. Increases in internalizing and externalizing symptoms from pre-pandemic to the initial lockdown period were found in Canada and the United Kingdom (Bignardi et al., 2021; Khoury et al., 2021). In a cohort of Norwegian children, an overall increase in mental health issues were reported during the early stages of the pandemic to 9 months after the initial outbreak, with internalizing symptoms accounting for the most increases (Lehmann et al., 2022). Factors such as parent stress, intra-family dysregulation, and screen time contributed to increases in children’s mental health problems (Khoury et al., 2021; Li & Zhou, 2021; Oliveira et al., 2021).
Excessive screen time use appears to have an adverse influence on children’s mental health outcomes (McArthur et al., 2022; Tamana et al., 2019; Tandon et al., 2021). In a large longitudinal study of brain development and child wellbeing in the United States (ABCD study; https://abcdstudy.org) increased screen time was associated with impaired social skills (Jericho & Elliott, 2020; Paulus et al., 2019), depression and anxiety (Fors & Barch, 2019), behavioural and social issues (Guerrero et al., 2019), and reduced sleep duration, fatigue, and insomnia in children (Hisler et al., 2020). Screen time was associated with a dose-dependent increase in depression risk in children (Liu et al., 2016). Youth reporting over 4 hours of passive screen time per day were significantly more likely to meet the criteria for major depressive episode, social phobia, and generalized anxiety disorder (Kim et al., 2020). By contrast, a systematic review (n = 159, 425) on children’s screen time behaviours found that screen time was weakly associated with externalizing and internalizing problems (Eirich et al., 2022). Age-related factors appear to influence the strength and direction of the association (Neville et al., 2021).
Adverse mental health outcomes associated with unregulated screen time is an emergent concern amid the ongoing pandemic. Cross sectional evidence from early stages of the pandemic demonstrates associations between excessive screen time and increased incidence of total mental health difficulties in children and youth, with sleep duration, physical activity and sedentary behaviour mediating these relationships (Olive et al., 2022; Tandon et al., 2021). In a cohort of 2026 Ontario children, more digital media and TV consumption was associated with conduct and hyperactivity/inattention problems in children under the age of 4 and internalizing problems in older children (Li et al., 2021).
Parent mental health, parenting strategies, and involvement with children’s academic and recreational activities may mediate mental health problems and excessive screen time use in children (Arundell et al., 2020; Barroso et al., 2018; Kochanova et al., 2022; Seguin et al., 2021; Tang et al., 2018). For example, under-resourced parents tend to perceive screen time as an effective tool to modify problem behaviour (Nikken, 2019). Greater maternal stress is associated with less screen-time monitoring/limit setting in children, while greater paternal stress is positively associated with limit setting (Tang et al., 2018).
The pandemic has contributed to heightened perceived stress among parents with school aged children (Adams et al., 2020; Calvano et al., 2021; Orgilés et al., 2020; Seguin et al., 2021), which may put children at risk for greater screen time use and mental health problems in the future (Hartshorne et al., 2020; Li & Zhou, 2021). For example, greater parent anxiety during the initial COVID-19 outbreak was a significant predictor of children’s internalizing and externalizing behaviours (Li & Zhou, 2021). In addition, over-reactive, inconsistent, and authoritarian parenting practices contributed to higher screen time use in children (Eyimaya & Irmak, 2021; Oliveira et al., 2021).
While studies point to the separate negative influences of screen time and parent stress on children’s mental health outcomes, few studies have addressed these concerns longitudinally. The influence of parental stress and parenting strategies has been identified as important mediating factors. It is essential to identify the factors that promote or impede childhood development, and to devise early intervention strategies to support parents and children during periods of school closures.
In the current longitudinal study, we examined the association between screen time, parenting styles, and parent stress on the incidences of externalizing, and internalizing behaviours in a cohort of Canadian children from November 2020 to November 2021. Two research questions were examined: 1) Is screen time associated with internalizing and externalizing behaviours in children over time? 2) Will changes in parent stress and parenting styles contribute to children’s mental health outcomes? We hypothesize that protracted screen time exposure will contribute to increased externalizing and internalizing behaviours, and that parent stress and parenting styles will moderate the strength of this association.
Parents (aged 18 years and over) and child (aged 6–12 years) residing in Canada with children receiving education through the public-school systems were invited to participate in this study. Participants were recruited online through Prolific and social media (i.e., Facebook, Twitter, and Instagram). Parents provided informed consent/assent. The study received approval from Western’s Non-Medical Research Ethics Board.
We gleaned data from an ongoing longitudinal community-based cohort project on children’s mental health and learning outcomes during the pandemic (COVID-19: Managing Parent Attitudes and School Stress). Data from two time points across the 2020-21 academic school year were considered for the present study. Baseline data were collected from parents between November 2020 and April 2021 on a rolling recruitment basis. Parents completed follow up surveys approximately 12 months from their baseline measures (June 2021- November 2021). Parents completed demographic measures and answered questions about their parenting styles, stress levels, and their child’s mental health and screen time behaviours. We condensed the demographics questionnaire at time two and incorporated supplementary items to capture pandemic-related alterations to school and home routines. All other procedures remained the same.
Parents provided information regarding their household income, employment status, age, gender, geographical location, and the number of children living at home. Children’s information, including age, special needs and mental health diagnoses, and mode of education, was provided by the participating parent.
Parenting styles were assessed using two subscales from the abbreviated Alabama Parenting Questionnaire (Frick, 1991). The APQ is a 28-item questionnaire that provides a score for parental involvement (how engaged the parent is in their child’s learning) and consistency with positive discipline. The Parent Stress Index -Fourth Edition Short Form (PSI-4; Cohen et al., 1983) was used to assess the magnitude of stress experienced by parent-child dyads. The PSI-4 contains three 12-item subscales, including Parental Distress, Parent-Child Dysfunctional Interaction, and Difficult Child. These subscales were summed to create the Total Stress Index.
Our primary interest was to examine recreational screen time, which was captured through parent estimates of time spent on social media, video gaming, and watching television. Parents provided estimates for sleep duration, time spent engaged in physical activity, and time spent on homework. Time-based estimates were measured in minutes per day. Parents indicated whether their child was enrolled in online or in-person schooling, as well as the parent’s satisfaction with their child's education and whether their child was engaged in their schoolwork.
We examined children's internalizing and externalizing behaviours through the Strengths and Difficulties Questionnaire (SDQ; Goodman, 1997). The SDQ is a validated behavioural questionnaire for caregivers of children (3–16 years old). The SDQ utilizes five subscales to assess emotional and behavioural development (i.e., emotional symptoms, peer relationship problems, conduct problems, hyperactivity–inattention, and prosocial behaviour). The sum of the conduct problems and hyperactivity-inattention scales yielded a total score for externalizing behaviours; the sum of peer relationship problems and emotional difficulties yielded a total score for internalizing behaviours. A Total Difficulties score was created by adding the composite scores for internalizing and externalizing behaviours.
All statistical analyses were performed using IBM SPSS Statistics software (Version 28, Statistical Package for the Social Sciences, IBM, Armonk, NY). A multiple imputation procedure addressed missing values in the dependent variables. We performed an analysis of missing values and Little’s MCAR test to determine whether missing data were missing completely at random (MCAR). Missing values were imputed using predictive mean matching. Our first aim was to examine whether screen time was associated with children’s mental health difficulties. Internalizing and externalizing behaviours were entered as dependent variables in separate generalized estimating equations models (GEE) with screen time entered as the independent variable. Time was included as a within-subjects factor to account for repeated measures. Our second aim was to determine whether parent stress and parenting styles modified the relationship between screen time and mental health outcomes. This was addressed by introducing parent stress and parental involvement into the models as covariates. All models were adjusted for family income, parent gender, child age, number of children living at home, sleep, physical activity, and special needs.
A total of 210 parents completed the questionnaire at T1. After excluding ineligible participants and those missing data, a total of 193 participants were retained for the study and 108 participants completed measures at the one year follow up (n = 113) prior to data cleaning). Participants who completed measures at both time points (51%, n = 108) were included in the final analysis. The median age was 38.00 (IQR = 8) years for parents and 8.00 (IQR = 3) years for children. Participant demographics are presented in Table 1.
The average time children spent on screens was 252.20 minutes (4.40 hours; SE = 18.45) at baseline and 233.99 at follow up (3.89 hours; SE = 16.70). No significant differences in daily screen time were evident between the two time points (t(103) = 1.01, 95% C.I. [-17.63 – 54.05], p= .316). No significant differences were found for internalizing (t(104) = .652, 95% C.I. [-.408 – .808], p = .516), or externalizing behaviours (t(104) = .096, 95% C.I. [-.561 – .619], p = .924) behaviours between the two time points.
Parent stress levels increased significantly from baseline (M= 78.45, SE= 1.89) to the second time point ([M= 86.23, SE= 2.59], (t(104) = -4.59, 95% C.I. [-11.13 – -4.41], p <.001)). No significant differences were found in parental involvement scores between time points (t(104)= .758, 95% C.I. [-.508 –1.14], p = .450).
To address our first aim, we examined the relationship between screen time and internalizing behaviours across the two time points. Internalizing behaviours were entered as the dependent variables and screen time was entered as a continuous predictor in the first step of the GEE. A significant positive main effect was found for screen time and internalizing behaviours (p= .030, Table 2: model 1). Physical activity (p = .001) and child age (p = .027) were negatively correlated with internalizing behaviours.
Parent stress and parental involvement were entered as covariates to the second step of the GEE for internalizing behaviours. Screen time and parent stress were positive predictors of internalizing behaviours (both, p <.05, Table 2: model 2). Parental involvement, and number of children at home were also positively associated with internalizing behaviours (both, p<0.05). Decreased levels of physical activity was negatively associated with internalizing behaviours.
An additional GEE was performed to examine the interaction effects between parent stress and screen time with internalizing behaviours (Table 2: model 3). Children who spent more time on screens had a parent who reported higher stress levels (p < .001, Fig. 1). Physical activity, number of children at home, and child age were also significant predictors in the model (all, p <.05).
Our second aim was to address whether screen time was associated with externalizing behaviours over time. No association was found between screen time and externalizing behaviours after adjusting for covariates in the GEE model (p>0.05). Parent stress and parental involvement were introduced to the model as moderator variables. Parent stress was a positive predictor of externalizing behaviours (β = 0.43, 95% C.I. .021-.066, p <.001). No other significant results were found.
The current study explored whether children with higher screen time use would exhibit more externalizing and internalizing behaviours than children with lower screen time use. We also considered the moderating influence of parental stress and involvement on behavioural outcomes and univariate changes to screen time, parenting factors and child mental health.
Children's screen time soared during the initial wave of the pandemic, with several North American and European countries reporting estimates of over 6.00 hours per day (Eyimaya & Irmak, 2021; Nagata et al., 2022; Seguin et al., 2021; Xiang et al., 2020). While we found no significant changes in screen duration over the study period, children were spending an average of 4.00 hours per day on screens, which remains higher than pre-pandemic levels and exceeds the recommended guidelines for Canadian children. This finding illustrates that continuous screen exposure over protracted periods may still pose health risks to children and adolescents.
Parent stress levels were within normal limits at both time points; however, there was a significant 7-point increase at the 1-year follow-up from baseline. This finding corresponds with, and extends upon, previous observations of parents' perceived stress amidst the pandemic. Previously, parents reported experiencing high-stress levels at the outset of the pandemic (Adams et al., 2020; Calvano et al., 2021; Seguin et al., 2021), with stress levels waning toward the end of 2020 (Adams et al., 2020). We found an increase in perceived stress toward the end of 2020 and into 2021, suggesting that parents continue to struggle with competing demands presented at home and in the workplace.
Children's internalizing behaviours were outcomes of interest in the current study. We found a moderate positive association between screen time and internalizing behaviours after controlling for demographic covariates, sleep duration, and physical activity. This observation corresponds with Eirich and colleagues’ (2022) meta-analysis demonstrating weak associations among screen time and children's internalizing and externalizing behaviours. Outcome heterogeneity was moderated by between-study variation.
Age-related factors could moderate the correlational strength of screen time and internalizing behaviours. For example, Neville and colleagues (2021) found that greater screen time among preschoolers significantly predicted internalizing behaviours at ages 5 and 7, but greater screen time at age 7 was significantly associated with less internalizing behaviors at 9 years of age. Screen time may disproportionately influence internalizing symptoms in children depending on their age (Li et al., 2021). Therefore, the modest association found in the current work may reflect the association between media use and internalizing symptoms in school-aged children. Nevertheless, screen time overexposure (≥ 4 hours per day) puts children at a threefold risk for developing major depressive disorder, social phobia, and generalized anxiety disorder (Kim et al., 2020), which underscores the necessity of limiting children's screen time activity, irrespective of age.
Parent stress and parental involvement modified the influence of screen time on internalizing behaviours. We also found evidence of a strong relationship between screen time and parent stress on the incidence of anxiety and depressive symptoms in children. Previous research demonstrates a strong positive association between parent mental health and screen time behaviours, which is mediated by dysfunctional parent-child interactions, inconsistent screen time monitoring (Tang et al., 2018), limited access to childcare resources (Hartshorne et al., 2020), lower household income (Lee et al., 2022; Seguin et al., 2021), extended school closures (Lee et al., 2022), and negative parenting strategies (Oliveira et al., 2021). Altogether, our findings indicate that children with greater screen time engagement and who dwell in high-stress environments are at significant risk of developing internalizing behaviours.
Our second outcome of interest was externalizing behaviours, which were comprised of hyperactivity/inattention symptoms and conduct problems. We found no association between screen duration and externalizing behaviours in school-aged children. Previous studies indicate that externalizing issues emerge more frequently in younger children. For example, Neville and colleagues (2021) found a directional association between externalizing behaviours and screen time activity in preschoolers, suggesting that parents may use technology to regulate challenging behaviour exhibited by young children. Spending over 2 hours per day on screens was found to increase preschoolers’ risk for developing inattention issues by the age of 5; however, the association with aggressive behaviours was not apparent (Tamana et al., 2019). Overall, screen time and externalizing difficulties appear weakly correlated (Eirich et al., 2022), suggesting that other latent factors may underpin children's behavioural functioning.
Parent stress emerged as a strong predictor for children's externalizing behaviours over the pandemic. The association between child behavioural difficulties, parent-child dysfunctional interactions, and parent stress is well established in the literature (Barroso et al., 2018; Kochanova et al., 2022; Oliveira et al., 2021). Parent stress is associated with externalizing behaviours in adolescents, which underscores the need to address parenting strategies and parent mental health when treating children with externalizing difficulties (Eirich et al., 2022; Kochanova et al., 2022).
A limitation of the current study was the use of online survey questionnaires to collect information. Albeit virtual collection methods attenuate pandemic-related bottlenecks presented in research, the lack of in-person observation exacerbates the risk of systemic and random error. A strength of our study was the compatibility with multiple technologies (i.e., tablets, smartphones, and laptops) and accessible to English and French speaking families. Another theoretical limitation of the current work is that children’s outcome data were exclusively derived from parent-report estimates. Children often have disparate perceptions of their school and home relationships, which may differ from parental reports. Finally, our study had modest attrition rates.
The current study provides insight to the factors that promoted or impeded children’s emotional and behavioural outcomes over the 2020–2021 academic school year; and the influence of parent stress and parenting. Screen time usage in children increased during the initial phases of the pandemic. Our current longitudinal study found elevated screen usage persisted in children following a year of pandemic life. Elevated screen use was associated with increased anxious and depressive symptoms in children, and parental stress. Future studies are needed to investigate the long-term effects of pandemic life on children's development.
Table 1 Parent demographics in the full cohort
Characteristic |
|
Full sample |
|
|||||
|
n |
% |
|
|||||
Parent Gender |
|
|
|
|
||||
Women |
|
56 |
52.8 |
|
||||
Men |
|
50 |
47.2 |
|
||||
Parent age groups |
|
|
|
|
||||
19–29 |
|
11 |
10.5 |
|
||||
30–39 |
|
56 |
53.3 |
|
||||
40–49 |
|
35 |
33.3 |
|
||||
50–59 |
|
3 |
2.9 |
|
||||
Primary parent a |
|
102 |
96.2 |
|
||||
Cohabitate a |
|
99 |
94.3 |
|
||||
Number of children at home |
|
|
|
|
||||
1–2 |
|
82 |
78.1 |
|
||||
3–4 |
|
21 |
20 |
|
||||
5+ |
|
2 |
1.9 |
|
||||
Income |
|
|
|
|
||||
Low < 40K |
|
7 |
7.1 |
|
||||
Moderate 40- 80K |
|
29 |
29.6 |
|
||||
High > 80K |
|
62 |
63.3 |
|
||||
Educational level |
|
|
|
|
||||
Highschool |
|
6 |
5.7 |
|
||||
College/some college |
|
22 |
21.0 |
|
||||
University/some university |
|
54 |
51.4 |
|
||||
Masters/postgraduate |
|
18 |
17.1 |
|
||||
Doctorate |
|
5 |
4.8 |
|
||||
Employment a |
|
93 |
12.3 |
|
||||
Spouse Employment a |
|
80 |
85.1 |
|
||||
Child age groups |
|
|
|
|
||||
6 – 7 |
|
39 |
39.4 |
|
||||
8 – 9 |
|
27 |
27.3 |
|
||||
10 – 12 |
|
33 |
33.3 |
|
||||
School delivery |
|
|
|
|
||||
Online |
|
40 |
37.7 |
|
||||
In-person |
|
66 |
62.3 |
|
||||
School enrollment |
|
|
|
|
||||
Full-time |
|
103 |
98.1 |
|
||||
Part-time |
|
2 |
1.9 |
|
||||
Child with special need a |
|
11 |
10.2 |
|
a Reflects the number and percentage of participants answering “yes” to this question
Table 2 Multivariable GEE regression model for children’s Internalizing behaviours
|
|
|
Model 1 |
|
|
|
|
Model 2 |
|
|
|
|
|
Model 3 |
|
|
|
|||||||||||||||||||||||
|
B |
SE |
95% CI |
p |
|
B |
SE |
95% |
CI |
p |
|
B |
SE |
95% |
CI |
p |
|
|||||||||||||||||||||||
LL |
UL |
|
|
|
|
LL |
UL |
|
|
|
|
LL |
UL |
|
||||||||||||||||||||||||||
Parent gender a |
-.590 |
.55 |
-1.67 |
.487 |
.283 |
|
-.577 |
.468 |
-1.49 |
.340 |
.218 |
|
-.548 |
5.30 |
-1.59 |
.490 |
.301 |
|||||||||||||||||||||||
Special needsb |
-1.94 |
1.00 |
-3.89 |
.015 |
.052 |
|
-2.24 |
.810 |
-3.82 |
-.648 |
.006* |
|
-2.06 |
.931 |
-3.89 |
-.237 |
.027* |
|||||||||||||||||||||||
Income |
.002 |
.007 |
-.011 |
.015 |
.738 |
|
.000 |
.006 |
-.011 |
.012 |
.959 |
|
.002 |
.006 |
-.011 |
.014 |
.768 |
|||||||||||||||||||||||
Children at home |
.450 |
.265 |
-.069 |
.970 |
.089 |
|
.479 |
.202 |
.084 |
.875 |
.017* |
|
.508 |
.245 |
.028 |
.988 |
.038* |
|||||||||||||||||||||||
Child age |
-.255 |
.115 |
-.481 |
-.029 |
.027* |
|
-.172 |
.100 |
-.361 |
.017 |
.075 |
|
-.265 |
1.06 |
-.472 |
-.059 |
.012* |
|||||||||||||||||||||||
Physical activity |
-.008 |
.002 |
-.012 |
-.003 |
.001** |
|
-.006 |
.002 |
-.010 |
-.002 |
.007* |
|
-.007 |
.002 |
-.012 |
-.003 |
.001** |
|||||||||||||||||||||||
Sleep |
.001 |
.004 |
-.007 |
.009 |
.806 |
|
.002 |
.003 |
-.005 |
.009 |
.570 |
|
.003 |
.004 |
-.005 |
.010 |
.482 |
|||||||||||||||||||||||
Screen time |
.003 |
.001 |
.000 |
.006 |
.030* |
|
.002 |
.001 |
.000 |
.005 |
.036* |
|
|
|
|
|
|
|||||||||||||||||||||||
PSIc |
|
|
|
|
|
|
.068 |
.010 |
.048 |
.088 |
.001** |
|
|
|
|
|
|
|||||||||||||||||||||||
Parental Involvement |
|
|
|
|
|
|
.086 |
.041 |
.006 |
.116 |
.034* |
|
.007 |
.035 |
-.062 |
.076 |
.840 |
|||||||||||||||||||||||
PSI*Screen time |
|
|
|
|
|
|
|
|
|
|
|
|
<0.001 |
1.45E-5 |
3.01E-5 |
8.71E-5 |
.001** |
a 0 = Female; 0 b = No developmental disorder; c Parent stress index. * p<.05, **p<.001