ABSTRACT
Aim
To compare physical activity and screen time behaviour of Swedish youth between three pandemic phases using data from an annual survey, and to evaluate the feasibility of device‐measured physical activity for future survey rounds.
Methods
A repeated cross‐sectional study with data from a Swedish population‐based survey (2018–2023) with questions on physical activity and screen time behaviour. The study period was categorised into prepandemic, pandemic and postpandemic phases. In 2022, screen time quality was included and a subsample used accelerometers (n = 700).
Results
A total of 50 163 children and adolescents (4–17 years) were included (49.0% girls, mean age 10.2 years). Regression models showed that physical activity was higher before (unstandardised B [95% CI] = 0.057 [0.002–0.112]) and after (0.046 [−0.004; 0.095]) the pandemic. Screen time was lower prepandemically (OR [95% CI] = 0.759 [0.726; 0.794]), with no significant difference during and after the pandemic. Data from 94% of the accelerometers were obtained, and these participants had higher socioeconomic status compared with the survey population.
Conclusion
Screen time increased during COVID‐19 and remained high after, while physical activity levels seemed less affected by the pandemic phases. Future studies would benefit from screen time quality assessment and complementary device‐based physical activity measurements.
Keywords: physical activity, screen usage, temporal trends, wearable devices
Summary.
This is the first study in Sweden evaluating trends in physical activity and screen time before, during and after the pandemic for 4‐ to 17‐year‐olds.
Screen time was higher during the pandemic and seemed to remain high postpandemically, while physical activity results were more difficult to interpret but tended to be lower during the pandemic before levelling off after the pandemic.
To better understand physical activity and screen time behaviour in Swedish youth, assessments of the types of screen time and device‐based physical activity assessment could be used.
Abbreviations
- App
Application
- LPA
Light physical activity
- MVPA
Moderate‐to‐vigorous physical activity
- PA
Physical activity
- WHO
World Health Organisation
1. Background
Physical activity in children and adolescents is associated with multiple beneficial health outcomes, such as better cardiorespiratory and musculoskeletal fitness, cardiometabolic health, bone health, and cognitive function, as well as academic outcomes [1, 2]. Additionally, there is a growing body of evidence that sedentary behaviour, especially recreational screen time, is related to poorer health outcomes [1, 3, 4].
Monitoring physical activity behaviour of children and adolescents over the years is an important source of information to assess current national trends and for policymakers and stakeholders when making decisions about youth health promotion. A pooled analysis of cross‐sectional survey data from 146 countries showed that the prevalence of physical inactivity remained concerningly high from 2001 to 2016, with more than 80% of adolescents still not meeting the physical activity recommendations in 2016 [5]. This detrimental trend was further exacerbated during the COVID‐19 pandemic [6], and concerns have been raised showing that the growing usage of screen time and social media among youth coincides with less physical activity and poorer health [7]. Therefore, new guidelines on physical activity and sedentary behaviour for children and adolescents have been published by the World Health Organisation (WHO) [1, 8].
During the COVID‐19 pandemic, external conditions for physical activity behaviour have been complicated by national restrictions. In Sweden, preschools, primary and secondary schools remained open, while classes in high schools and universities were held via distance learning [9]. In addition, physical distancing and limiting social gatherings might have influenced children's health behaviour. In a small study, preschool‐aged children in Sweden were actually found to have an increased level of physical activity [10], while larger studies including older Swedish participants found either no change in physical activity [11, 12] or only a change in light but not moderate‐to‐vigorous physical activity (MVPA) [13]. All studies consistently reported an increase in screen time [10, 11, 12, 13]. However, little is known about how these trends continued in the postpandemic phase, after most of the Swedish restrictions were suspended in February 2022 [14]. With this background, our aim was to compare physical activity and screen time behaviour of Swedish children and adolescents between the prepandemic, the pandemic and the postpandemic years using data from an annually repeated population‐based survey (the Generation Pep study [15]). Also, the Generation Pep surveys from previous years only measured physical activity levels by means of subjectively reported data and did only assess screen time quantity; therefore, as secondary aims, we wanted to (A) investigate the feasibility of adding device‐measured physical activity to the Generation Pep survey, which could provide complementary information in the future and (B) provide further details on screen time usage.
2. Methods
2.1. Study Design and Participants
The Generation Pep study is based on an annually repeated cross‐sectional data collection in Swedish children and adolescents aged 4–17 years. The survey has been conducted over the last 6 years (2018–2023). Every year, a random probability sample of 29 000 children and adolescents is drawn from the Swedish population registry, and invitations to participate are sent out to the registered postal addresses for the child's caregiver in late August or the beginning of September of each year. Data collection is terminated by the end of the year. The material that is sent out includes information regarding the study, a unique username and password and an anonymous link to the web‐based questionnaire. To increase the participation rate, one postal as well as three to four SMS reminders are sent to the families who have not completed the questionnaire. Informed consent is obtained from the caregiver participating with the child/adolescent as the first part of the web‐based questionnaire.
The survey includes around 70 questions and is divided into two parts: the first one addresses the caregivers, while the second one is directed towards the children or adolescents. The questions for the caregivers are about their sociodemographic characteristics, about their current living situations, about the demographic characteristics of their child and medical issues, as well as the special forms of diet of the child. The second part of the questionnaire is answered either by the children themselves (≥ 12 years) or by the caregiver and the child together (< 12 years). A detailed description of the population sampling and the study flow has been published previously [11, 15].
This study has been approved by the Swedish Ethical Review Authority (2018/375–31/5; 2018/1461–32; 2019–03798; 2020–03853; 2021–03931; 2022–03926‐02; and 2023–04417‐02). The participants who completed the questionnaire received a small gift to thank them for their participation (e.g., a gift certificate for ≈100 SEK, a 1‐month subscription to a music streaming service, etc.).
2.2. Pandemic Phases
We categorised the observation time into three pandemic phases. We defined the Years 2020 and 2021 as the pandemic phase because the pandemic started in early 2020, and in February 2022, most of the restrictions were lifted in Sweden [14]. Consequently, the Years 2018–2019 were classified as ‘pre‐pandemic’ and 2022–2023 as ‘post‐pandemic’.
2.3. Outcomes
2.3.1. Physical Activity Questions From the Survey
There are two questions assessing physical activity in the questionnaire: The first question is as follows: ‘Think about yesterday, approximately how long were you physically active during the day?’. The participants had to choose from a 7‐point rating scale ranging from 1 (no physical activity) to 7 (between 2.5 and 3 h of physical activity).
The second question is as follows: ‘Think about the past week – that is, the last seven days from today – how many days have you been physically active for a total of 60 min or more?’. The participants had to choose between 1 and 7 days. Both questions were accompanied by the explanatory sentence, ‘Any activity that makes your heart beat faster and sometimes leaves you short of breath counts’.
In 2019, the Generation Pep team updated the survey question on physical activity yesterday. Instead of a continuous scale, an ordinal scale was implemented for a subsample of the 2019 cohort and the complete cohorts of the following years. Consequently, for harmonisation, in this evaluation, we compare trends in physical activity between the Years 2019 (subsample) and 2023.
2.3.2. Screen Time Questions From the Survey
Screentime was assessed using the following question: ‘Think about yesterday, approximately how long in total have you been sedentary in front of a screen outside of class time, i.e., mobile phone, television, computer screen, tablet?’. The participants had to choose their answer from an ordinal scale from 1 (‘not at all’) to 6 (‘more than 7 hours’). The screen time question was not included in the 2023 survey; hence, screentime trends were evaluated for the Years 2018–2022.
In 2022, additional questions about the types of screen time were asked as part of a special focus investigation that year. The participants were asked, ‘You answered a previous question regarding your time in front of a screen yesterday. Of the total time you spent in front of a screen yesterday, approximately how many minutes did you spend on the following activities?’; (a) social media (Instagram, Snapchat, TikTok, etc.), (b) gaming (mobile games, computer games or video games), (c) watching movies, series, clips or live streams (TV, Netflix, YouTube, Twitch, etc.), (d) reading (news, articles, e‐books, etc.), (e) schoolwork (homework or digital learning material), (f) creating own content (drawing, video/photo editing, music production, etc.), and (g) other (free text option). To improve accuracy of our analyses, we calculated the total screen time yesterday from this question and compared it to the categorical screen time question described above. For the types of screen time analysis, we only used data from participants with a consistent answer in both questions (+/−1 h).
2.3.3. Device‐Measured Physical Activity
In 2022, all participants who were invited to respond to the Generation Pep survey were invited to participate in an additional investigation with the aim of assessing the feasibility of using a device‐based assessment of physical activity behaviours. The specific feasibility outcomes investigated were participation rate, available and valid data and representativeness of the population.
The procedure of data collection for this substudy was as follows: All individuals invited to the survey were informed that a maximum of 700 children could attend the device‐measured physical activity evaluation, and 700 signed up shortly after receiving the invitation. After caregivers had given their consent, accelerometers and instructions on how to attach them to the children's outer side of the thigh were sent out to the families. We used the SENS motion activity sensor (SENS Innovation ApS, Copenhagen, Denmark), which has been previously validated in healthy children [16]. The sensor is a three‐axis accelerometer with a sampling frequency of 12.5 Hz [16]. Data are analysed in 5‐s epochs, and each epoch is estimated to belong to a specific activity category (see Table S1) according to the SENS algorithm (SENS motion User Manual). The participants were instructed to wear the accelerometer for 7 full consecutive days for 24 h each day, and to fill out a sleep log used to define sleep time in the analyses. Thereafter, the device and sleep log were sent back to the research team, where raw accelerometry data were transferred from the SENS motion device to the SENS motion mobile application (app). A valid day was defined as 16 h of wear time, including sleep and time awake. Participants included in the analyses had filled out the sleep log and had at least one valid day of wear time.
2.4. Statistical Analyses
Descriptive characteristics were presented as number of observations and percentages for categorical variables and as means and standard deviations for ordinal and continuous variables. The associations between the ordinal questions (physical activity yesterday and screen time yesterday) and the categorical years were calculated using an ordered logistic regression model, reporting the odds ratio of being in a higher rather than in a lower answer category than in the reference year. The association between the ordinal physical activity question (days with at least 60 min of physical activity last week) was assessed with linear regression models reporting unstandardised B coefficients. As reference year, we used the first year of the survey results. For the association between pandemic phases and questionnaire data, we used the same approach as described above and categorised the years. As reference, we used the pandemic phase from 2020 to 2021 and compared it to the pre‐ and postpandemic phases. All calculations were performed unadjusted (Model 1) and adjusted for children's age, sex as well as caregiver's education status and household income (Model 2). We used bar charts to visualise categorical and continuous descriptive data, and a forest plot for the comparison of the pandemic phases.
Analysis was performed using STATA version 18.0 (StataCorp LP, College Station, Texas, USA), and p < 0.05 was considered statistically significant.
3. Results
Overall, 50 163 Swedish children and adolescents participated in the survey from 2018 to 2023, of which 49.0% were girls, and the mean age was 10.2 years (Table 1). Although the net participation rate, calculated as having answered at least one question, somewhat declined, the baseline characteristics of the children remained robust over the years. Respondents living in a large city, the percentage of primary caregivers with a college degree and parents with a household median population income, from Statistics Sweden, ≥ 40 000 SEK increased over time.
TABLE 1.
Baseline characteristics stratified by years.
| Year | |||||||
|---|---|---|---|---|---|---|---|
| 2018 | 2019 | 2020 | 2021 | 2022 | 2023 | Total | |
| N | 12 657 (25.2%) | 9035 (18.0%) | 8320 (16.6%) | 7045 (14.0%) | 6039 (12.0%) | 7067 (14.1%) | 50 163 (100.0%) |
| Response rate | 43.6% | 31.2% | 28.7% | 24.3% | 20.8% | 24.4% | 34.6% |
| Children's sex a | |||||||
| Girl | 5993 (49.2%) | 4109 (48.5%) | 3876 (49.7%) | 3178 (48.7%) | 2709 (49.0%) | 3137 (48.8%) | 23 002 (49.0%) |
| Boy | 6136 (50.4%) | 4319 (51.0%) | 3887 (49.8%) | 3312 (50.8%) | 2776 (50.3%) | 3246 (50.5%) | 23 676 (50.4%) |
| Other | 19 (0.2%) | 18 (0.2%) | 19 (0.2%) | 14 (0.2%) | 23 (0.4%) | 20 (0.3%) | 113 (0.2%) |
| Not willing to specify | 33 (0.3%) | 28 (0.3%) | 24 (0.3%) | 16 (0.2%) | 15 (0.3%) | 29 (0.5%) | 145 (0.3%) |
| Children's age b , mean (SD) | 10.236 (3.863) | 10.098 (3.827) | 10.100 (3.858) | 10.082 (3.866) | 10.229 (3.854) | 10.216 (3.901) | 10.163 (3.860) |
| Place of residence | |||||||
| Large cities | 4744 (37.5%) | 3421 (37.9%) | 3302 (39.7%) | 2792 (39.6%) | 2309 (38.2%) | 2754 (39.0%) | 19 322 (38.5%) |
| Medium‐sized towns | 4857 (38.4%) | 3477 (38.5%) | 3113 (37.4%) | 2670 (37.9%) | 2364 (39.1%) | 2789 (39.5%) | 19 270 (38.4%) |
| Smaller towns/urban areas | 3056 (24.1%) | 2137 (23.7%) | 1905 (22.9%) | 1583 (22.5%) | 1366 (22.6%) | 1524 (21.6%) | 11 571 (23.1%) |
| Primary caregiver with college degree c (> 3 years) | |||||||
| No | 5341 (43.3%) | 3246 (37.9%) | 2898 (36.8%) | 2322 (35.0%) | 2034 (36.1%) | 2316 (34.8%) | 18 157 (38.1%) |
| Yes | 6980 (56.7%) | 5328 (62.1%) | 4986 (63.2%) | 4312 (65.0%) | 3600 (63.9%) | 4342 (65.2%) | 29 548 (61.9%) |
| Household income above 40 000 SEK d | |||||||
| No | 1639 (15.7%) | 968 (13.2%) | 843 (12.5%) | 629 (11.2%) | 566 (11.8%) | 574 (10.1%) | 5219 (12.8%) |
| Yes | 8817 (84.3%) | 6361 (86.8%) | 5880 (87.5%) | 5008 (88.8%) | 4243 (88.2%) | 5126 (89.9%) | 35 435 (87.2%) |
Note: The total N is the number of responders who answered at least one question. All other percentages provided in the table refer to responders to the corresponding question.
Sex, nonresponse 2018 n = 476; 2019 n = 561; 2020 n = 514; 2021 n = 525; 2022 n = 516; 2023 n = 635; total n = 3227.
Age, nonresponse 2018 n = 496; 2019 n = 576; 2020 n = 525; 2021 n = 544; 2022 n = 535; 2023 n = 650; total n = 3326.
Education, nonresponse 2018 n = 336; 2019 n = 461; 2020 n = 436; 2021 n = 411; 2022 n = 405; 2023 n = 409; total n = 2458.
Income, nonresponse 2018 n = 2201; 2019 n = 1706; 2020 n = 1597; 2021 n = 1408; 2022 n = 1230; 2023 n = 1367; total n = 9509.
3.1. Trends in Physical Activity From 2019 to 2023
Figures 1 and 2 show the distribution of the reported physical activity over the years. According to both corresponding survey questions, the lowest physical activity was reported in the Year 2021 and the highest in 2023.
FIGURE 1.
Trends of responses about physical activity yesterday from 2019 to 2023. Only a subsample (n = 2726) is included in the physical activity analysis in 2019. Question: ‘Think about yesterday, approximately how long were you physically active during the day?’; answer: ordinal (seven categories); and results: yearly proportion of answers.
FIGURE 2.
Trends in the number of days with at least 60 min of physical activity last week. Question: ‘How many days have you been active for more than 60 min during the last week?’; answer: discrete (days, from 0 to 7); results: yearly means; and reference line: overall mean from 2019 to 2023.
3.1.1. Question 1: Physical Activity Yesterday
Overall, we found no significant changes in physical activity levels comparing the pandemic phase with the pre‐ and postpandemic phase in the results for Question 1 (Table 2). After stratification into age groups, there was a significantly lower level of physical activity during the prepandemic phase compared to the pandemic phase in 4‐ to 6‐year‐old children (adjusted OR 0.838, 95% CI 0.712–0.986, p = 0.033), while the opposite was found for physical activity levels in adolescents between 13 and 17 years (adjusted OR 1.183, 95% CI 1.022–1.370, p = 0.025). There were no significant changes between the pandemic and the postpandemic phase. In the regression models (Table S2), comparing the first year of the survey 2019, with the other years, we found no significant differences in the answers to categorical Question 1.
TABLE 2.
Association between physical activity and pandemic phases (physical activity yesterday).
| Model 1 | Model 2 | |||||
|---|---|---|---|---|---|---|
| OR | 95% CI | p | OR | 95% CI | p | |
| Overall | ||||||
| Prepandemic | 1.029 | 0.958–1.106 | 0.433 | 1.028 | 0.951–1.111 | 0.490 |
| Pandemic | 1.000 | 1.000 | ||||
| Postpandemic | 1.041 | 0.996–1.088 | 0.073 | 1.031 | 0.983–1.081 | 0.207 |
| Children's age | 0.953 | 0.947–0.959 | < 0.001 | |||
| Children's sex | 1.308 | 1.252–1.366 | < 0.001 | |||
| Primary caregiver with college degree (more than 3 years) | 1.168 | 1.112–1.227 | < 0.001 | |||
| Household income above 40 000 SEK | 1.516 | 1.406–1.636 | < 0.001 | |||
| Stratified by age categories | ||||||
| 4–6 years | ||||||
| Prepandemic | 0.843 | 0.721–0.985 | 0.032 | 0.838 | 0.712–0.986 | 0.033 |
| Pandemic | 1.000 | 1.000 | ||||
| Postpandemic | 1.001 | 0.912–1.098 | 0.991 | 0.992 | 0.900–1.093 | 0.873 |
| 7–12 years | ||||||
| Prepandemic | 1.048 | 0.944–1.164 | 0.375 | 1.033 | 0.923–1.155 | 0.572 |
| Pandemic | 1.000 | 1.000 | ||||
| Postpandemic | 1.065 | 0.998–1.137 | 0.056 | 1.024 | 0.955–1.097 | 0.512 |
| 13–17 years | ||||||
| Prepandemic | 1.138 | 0.998–1.297 | 0.053 | 1.183 | 1.022–1.370 | 0.025 |
| Pandemic | 1.000 | 1.000 | ||||
| Postpandemic | 1.051 | 0.971–1.138 | 0.216 | 1.067 | 0.977–1.166 | 0.149 |
| Stratified by sex | ||||||
| Girls | ||||||
| Prepandemic | 1.087 | 0.980–1.205 | 0.114 | 1.106 | 0.988–1.237 | 0.079 |
| Pandemic | 1.000 | 1.000 | ||||
| Postpandemic | 1.029 | 0.966–1.096 | 0.372 | 1.020 | 0.953–1.092 | 0.562 |
| Boys | ||||||
| Prepandemic | 0.979 | 0.884–1.083 | 0.678 | 0.962 | 0.862–1.073 | 0.486 |
| Pandemic | 1.000 | 1.000 | ||||
| Postpandemic | 1.064 | 1.000–1.132 | 0.051 | 1.050 | 0.982–1.123 | 0.157 |
Note: Question 1: ‘Think about yesterday, approximately how long were you physically active during the day?’ Answer: ordinal (seven categories): models: ordered logistic regression: results: odds ratios. (Note: In the ordered logistic case, it is the ratio, given a one‐unit increase in the covariate, of the odds of being in a higher rather than a lower category.) Model 1: unadjusted and Model 2: adjusted for age (except stratified analysis for age), sex of the participant (except stratified analysis for sex) and for education and household income of the caregiver.
3.1.2. Question 2: Days With at Least 60 Min of Physical Activity Last Week
Compared to the pandemic phase from 2020 to 2021, the physical activity level last week was significantly higher in the prepandemic phase (adjusted coefficient 0.057 days, 95% CI 0.002–0.112, p = 0.043) and tended to be higher in the postpandemic phase (adjusted coefficient 0.046 days, 95% CI −0.004 to 0.095, p = 0.071) (Figure 3a and Table 3). This association was mainly seen in adolescents aged 13–17 years, whereas no significant changes were shown for 4‐ to 12‐year‐old children (Table 3). After stratification for sex, we found that physical activity levels were significantly lower during compared to before the pandemic in girls, while there was no significant difference in the boys group and no significant changes comparing pandemic and postpandemic periods. By comparing each year with the initial year, we found a significantly lower mean number of days with at least 60 min of physical activity in 2021 compared to 2019 (adjusted coefficient −0.097 days, 95% CI −0.163 to −0.031, p = 0.004) (Table S2).
FIGURE 3.
Association between pandemic phases and physical activity/screen time (reference lines correspond to pandemic phase from 2020 to 2021). Physical activity: question: ‘How many days have you been active for more than 60 min during the last week?’; answer: discrete (days); models: linear regression models; results: coefficients. Screen time: question: ‘Think about yesterday, approximately how long in total have you been sedentary in front of a screen outside of class time, i.e., mobile phone, television, computer screen, tablet?’; answer: ordinal (categories); models: ordered logistic regression; and results: odds ratios. Model 1: unadjusted. Model 2: adjusted for age, sex of the participant and for education and household income of the caregiver.
TABLE 3.
Association between physical activity and pandemic phases (days with > 60 min of physical activity).
| Model 1 | Model 2 | |||||
|---|---|---|---|---|---|---|
| B | 95% CI | p | B | 95% CI | p | |
| Overall | ||||||
| Prepandemic | 0.048 | −0.005 to 0.101 | 0.074 | 0.057 | 0.002–0.112 | 0.043 |
| Pandemic | 0.000 | 0.000 | ||||
| Postpandemic | 0.041 | −0.007 to 0.089 | 0.093 | 0.046 | −0.004 to 0.095 | 0.071 |
| Children's age | −0.099 | −0.104 to −0.093 | < 0.001 | |||
| Children's sex | 0.285 | 0.243–0.326 | < 0.001 | |||
| Primary caregiver with college degree (more than 3 years) | 0.331 | 0.284–0.378 | < 0.001 | |||
| Household income above 40 000 SEK | 0.457 | 0.387–0.528 | < 0.001 | |||
| Stratified by age categories | ||||||
| 4–6 years | ||||||
| Prepandemic | −0.001 | −0.115 to 0.112 | 0.984 | 0.008 | −0.109 to 0.126 | 0.888 |
| Pandemic | 0.000 | 0.000 | ||||
| Postpandemic | 0.035 | −0.068 to 0.138 | 0.505 | 0.033 | −0.074 to 0.139 | 0.547 |
| 7–12 years | ||||||
| Prepandemic | 0.023 | −0.050 to 0.096 | 0.532 | 0.019 | −0.058 to 0.095 | 0.627 |
| Pandemic | 0.000 | 0.000 | ||||
| Postpandemic | 0.044 | −0.022 to 0.111 | 0.190 | 0.012 | −0.058 to 0.082 | 0.736 |
| 13–17 years | ||||||
| Prepandemic | 0.123 | 0.026–0.219 | 0.013 | 0.163 | 0.057–0.269 | 0.003 |
| Pandemic | 0.000 | 0.000 | ||||
| Postpandemic | 0.073 | −0.014 to 0.159 | 0.099 | 0.092 | −0.002 to 0.187 | 0.056 |
| Stratified by sex | ||||||
| Girls | ||||||
| Prepandemic | 0.088 | 0.013–0.163 | 0.022 | 0.090 | 0.012–0.169 | 0.024 |
| Pandemic | 0.000 | 0.000 | ||||
| Postpandemic | 0.029 | −0.039 to 0.097 | 0.408 | 0.036 | −0.034 to 0.107 | 0.314 |
| Boys | ||||||
| Prepandemic | 0.012 | −0.062 to 0.086 | 0.744 | 0.031 | −0.046 to 0.108 | 0.430 |
| Pandemic | 0.000 | 0.000 | ||||
| Postpandemic | 0.061 | −0.006 to 0.128 | 0.074 | 0.061 | −0.009 to 0.131 | 0.086 |
Note: Question 2: ‘How many days have you been active for more than 60 min during the last week?’, Answer: discrete (days from 0 to 7), Models: linear regression models, and Results: unstandardised coefficients (B). Model 1: unadjusted. Model 2: adjusted for age (except stratified analysis for age), sex of the participant (except stratified analysis for sex) and for education and household income of the caregiver.
3.2. Trends in Screen Time From 2018 to 2022
Figure 4 depicts the distribution of the answers regarding screen time over the years. The peak was in 2020, with 84.2% of the participants reporting more than 1 h of screen time yesterday, while in 2018, this percentage was at 79.4%.
FIGURE 4.
Trends in self‐reported screen time outside of school during the last day. Question: ‘Think about yesterday, approximately how long in total have you been sedentary in front of a screen outside of class time, i.e., mobile phone, television, computer screen, tablet?’; answer: ordinal (six categories); and results: yearly proportions of answers. N = 37 904 (Years 2018–2022).
Comparing the pandemic phase with the prepandemic phase, we found a significantly lower screen time in the prepandemic phase (adjusted OR 0.759, 95% CI 0.726–0.794, p < 0.001), but no significant difference between the pandemic and the postpandemic phase (Figure 3b and Table 4). Moreover, we found a significant increase in screen time with every year from 2018 to 2022. The largest increase was recorded for the Year 2020 (adjusted OR 1.438, 95% CI 1.356–1.526, p < 0.001) (Table S3). After stratification for age categories, the results throughout the pandemic phases remained robust for children from 7 to 17 years, but there was a decrease in screen time in the postpandemic phase in children from 4 to 6 years (Table 4). Similarly, after stratification for sex, we saw that the boys showed a significant decrease in screen time postpandemically.
TABLE 4.
Association between screen time and pandemic phases.
| Model 1 | Model 2 | |||||
|---|---|---|---|---|---|---|
| OR/Coefficient | 95% CI | p | OR/Coefficient | 95% CI | p | |
| Overall | ||||||
| Prepandemic | 0.799 | 0.768–0.832 | < 0.001 | 0.759 | 0.726–0.794 | < 0.001 |
| Pandemic | 1.000 | 1.000 | ||||
| Postpandemic | 0.982 | 0.926–1.042 | 0.553 | 0.950 | 0.891–1.014 | 0.124 |
| Children's age | 1.297 | 1.289–1.305 | < 0.001 | |||
| Children's sex | 1.274 | 1.225–1.326 | < 0.001 | |||
| Primary caregiver with college degree (more than 3 years) | 0.920 | 0.881–0.961 | < 0.001 | |||
| Household income above 40 000 SEK | 1.171 | 1.098–1.249 | < 0.001 | |||
| Stratified by age categories | ||||||
| 4–6 years | ||||||
| Prepandemic | 0.810 | 0.742–0.885 | < 0.001 | 0.797 | 0.726–0.874 | < 0.001 |
| Pandemic | 1.000 | 1.000 | ||||
| Postpandemic | 0.884 | 0.775–1.008 | 0.065 | 0.858 | 0.748–0.985 | 0.030 |
| 7–12 years | ||||||
| Prepandemic | 0.759 | 0.715–0.806 | < 0.001 | 0.774 | 0.726–0.826 | < 0.001 |
| Pandemic | 1.000 | 1.000 | ||||
| Postpandemic | 0.912 | 0.836–0.995 | 0.038 | 0.946 | 0.861–1.038 | 0.240 |
| 13–17 years | ||||||
| Prepandemic | 0.713 | 0.663–0.767 | < 0.001 | 0.718 | 0.662–0.779 | < 0.001 |
| Pandemic | 1.000 | 1.000 | ||||
| Postpandemic | 1.073 | 0.965–1.193 | 0.190 | 1.084 | 0.963–1.222 | 0.183 |
| Stratified by sex | ||||||
| Girls | ||||||
| Prepandemic | 0.772 | 0.729–0.818 | 0.000 | 0.742 | 0.697–0.791 | < 0.001 |
| Pandemic | 1.000 | 1.000 | ||||
| Postpandemic | 1.042 | 0.958–1.134 | 0.341 | 1.066 | 0.971–1.170 | 0.180 |
| Boys | ||||||
| Prepandemic | 0.832 | 0.786–0.881 | 0.000 | 0.781 | 0.733–0.831 | < 0.001 |
| Pandemic | 1.000 | 1.000 | ||||
| Postpandemic | 0.915 | 0.842–0.995 | 0.037 | 0.852 | 0.778–0.933 | 0.001 |
Note: Question: ‘Think about yesterday, approximately how long in total have you been sedentary in front of a screen outside of class time, i.e., mobile phone, television, computer screen, tablet?’. Model 1: Unadjusted. Model 2: Adjusted for age (except stratified analysis for age), sex of the participant (except stratified analysis for sex) and for education and household income of the caregiver.
3.3. Additional Analyses
3.3.1. Device‐Measured Physical Activity Data From 2022
From the 700 accelerometers that were sent to a subpopulation of the survey in 2022, data were obtained from 657 accelerometers, resulting in a participation rate of 94% (Figure S1). For our analysis, we used only complete data, which was available from 435 study participants (missing data mostly due to missing sleep logs).
The accelerometer cohort differed in some baseline characteristics from the overall cohort, especially regarding a higher socioeconomic background (Table S4). Overall, the children and adolescents spent on average 5 h of their day with physical activity of any intensity (light physical activity: 4.4 h, MVPA: 37 min) and 8.9 h with sedentary behaviour (Table S5). Adolescents > 12 years of age had a significantly lower number of overall physical activity, that is, light‐ and MVPA (coefficient −0.743 h [−1.144; −0.341]) and a higher amount of sedentary behaviour (coefficient 1.3 h [0.901; 1.711]) compared to 4‐ to 6‐year‐old children.
Children 7−12 years had significantly higher MVPA than children 4–6 years (coefficient 0.126 h [0.005; 0.206]), but no differences in overall PA or sedentary time between were found. Boys spent more time in MVPA than girls (coefficient 0.176 h [0.124; 0.228]), but there were no significant differences between sexes in sedentary behaviour and overall physical activity. Moreover, children with overweight or obesity had lower levels of MVPA compared with children of normal weight (coefficient −0.077 h [−0.150; −0.005]) (Table S6). Participants with parents with a low household income had a significantly higher sedentary behaviour compared to those with parents with average (coefficient −0.703 h [−1.290; −0.116]) or high incomes (coefficient −0.943 h [−1.514; −0.371]).
3.3.2. Types of Screen Time Analyses From 2022
Of the 4919 participants who reported any screen time on the day before the survey, 3597 (73.1%) had answers that were consistent with the question about types of screen time (+/−1 h). Overall, the most common screen time use was ‘Watching films, clips, etc.’, followed by ‘Gaming’ and ‘Social media’ (Figure 5A). After stratification into age classes, the results show that younger children from 4 to 6 years spend most of their screen time ‘Watching films, clips etc’ and ‘Gaming’, while ‘Social media’ becomes the most common type of screen time use in adolescents aged 13–17 years (Figure 5B). The main difference between girls and boys was that girls had more social media time, while boys were gaming more often (Figure 5C).
FIGURE 5.
Types of screen time. Question: ‘You answered a previous question regarding your time in front of a screen yesterday. Of the total time you spent in front of a screen yesterday, approximately how many minutes did you spend on the following activities?’; answer: ordinal (six categories); and results: yearly proportions of answers. N = 3743 (Year 2022).
4. Discussion
In this population‐based repeated cross‐sectional study, we observed physical activity and screen time behaviour of Swedish children and adolescents over the different pandemic phases. Hereby, we made two main observations. First, there was little to no impact of the COVID‐19 pandemic on the physical activity levels in our study cohort. And if there was, the lower physical activity levels during the pandemic years seemed to recover in the postpandemic phase. Second, compared to the prepandemic years, screen time during COVID‐19 was significantly higher and, in contrast to physical activity, there were no signs of less screen time postpandemically.
4.1. Trends of Physical Activity and Screen Time Over the Pandemic Phases
4.1.1. Prepandemic Versus Pandemic Years
Looking at the comparison between prepandemic and pandemic years, we found no significant difference overall when we analysed the question about physical activity the day before. Nevertheless, the number of days with at least 60 min of physical activity during the last week showed a significantly higher physical activity level in the prepandemic phase. After stratifying the results for both questions, we found that younger children (4–7 years) had in fact higher physical activity levels during the pandemic, while it was mostly the adolescents (13–17 years) who showed a lower physical activity level. Similar results were described by two other Swedish cohort studies; Nyström et al. investigated preschoolers from 0 to 5 years and found that during the pandemic, their physical activity levels were higher, possibly due to that preschools in Sweden were open during the COVID‐19 pandemic and that more time was spent outside [10]. Using device‐measured physical activity, Helgadóttir et al. showed a decrease in light physical activity (LPA) in 13‐ to 14‐year‐olds during the pandemic [13]. However, another study reported that most Swedish 13‐year‐olds maintained their levels of PA during the pandemic (10). Our data clearly showed higher amounts of screen time during the pandemic, and this was also found in all other Swedish studies mentioned above [10, 12, 13], as well as in an international meta‐analysis [17]. Interestingly, a Danish study in 6‐ to 11‐year‐old children from 2019/2020 found very similar percentages of excessive screen time users (13% of the participants had > 4 h/day) to what we found in our study (10% in 2019 and 13% in 2020) [18].
4.1.2. Pandemic Versus Postpandemic Years
Comparing the pandemic with the postpandemic phase, there was a trend towards higher physical activity levels after the pandemic, at least in one of the two survey questions. A research group from the United Kingdom, which has done more in‐depth analysis using device‐measured physical activity, has shown that after an initial drop, children's MVPA returned to prepandemic levels by July 2022, while sedentary time remained higher [19]. Another study describes a new postlockdown profile characterised by higher LPA, a higher percentage of children in the least active profile and a widening sex and socioeconomic gap [20]. In our cohort, we had no trend data for sedentary behaviour, but subjective data about screen time could be seen as a proxy. We did not find any differences between the pandemic and the postpandemic phase regarding screen time in the children and adolescents from 7 to 17 years and for girls, which indicates that within these groups, the amount of screen time remained higher compared to the prepandemic phase.
4.2. Evaluation of Secondary Aims
4.2.1. Device‐Measured Physical Activity
Within the context of this repeated survey on Swedish children and adolescents, our findings show that it is feasible to add device‐measured PA in a subcohort of the study. Firstly, the participation rate was high, with a 6% drop‐out rate indicating a large interest in providing such data in a population‐based sample. Furthermore, even though valid accelerometry data were only available for 62% of the participants, it is relevant to note that, if the missing sleep logs (20%, n = 143) had been collected, 82% of the participants could have been included in the analyses. This percentage is more in line with previous, similar studies [21, 22].
Compared with studies on device‐measured PA in children from the Nordic countries, the children in our study had somewhat lower mean daily MVPA levels (on average approximately 10 min per day) [18, 22, 23]. These findings are likely related to diverse assessment procedures in terms of different devices, placements and data procedures—a common challenge when comparing findings from device‐measured PA [23]. However, it is also possible that the lower MVPA levels in our study were reflective of less time spent in MVPA in 2022 compared to studies including data from the prepandemic years [21, 22]. Further, for sedentary time and overall PA, no sex differences could be detected, which contrasts with previous findings [18, 22]. Again, methodological differences, using different devices, placements and algorithms are likely explanations for the diverse findings. Another contrast to previous literature is the lack of difference in sedentary time between children in the youngest age groups, which might be reflective of the fact that the SENS motion accelerometer has not been validated in 4‐ to 6‐year‐old children, together with a small sample size for the youngest ages.
4.2.2. Types of Screen Time
Our findings show that the types of screen time for children 4–6 years mostly involved watching films and clips, and that social media use was more common among 13‐ to 7‐year‐olds and girls, which is in line with previous findings in children and adolescents [24, 25]. Interestingly, the subgroups mentioned to have the highest social media consumption in our cohort were also the ones with consistently higher overall screen time postpandemically, indicating that social media use might play an important role in this trend. This type of screen use has also been found to be positively associated with attention issues and conduct problems in children [26] and with depressive symptoms in adolescents [24, 27]. Moreover, the different types of screen time and not just screen time quantity seem related to mental health outcomes in adolescents and language skills in younger children [27, 28].
4.3. Implications for the Future
4.3.1. Evaluation of Screen Time
Based on our findings of more screen time during the pandemic and postpandemic years compared with before COVID‐19, it seems crucial to further monitor this behaviour, especially in children from 7 to 17 years. While traditionally, research of movement behaviour as well as international guidelines had focused more on physical activity rather than sedentary behaviour or screen time, this has changed in recent years. However, there are still a lot of unanswered questions. A review of the evidence, which was used for the WHO 2020 guidelines, highlights that there is sufficient evidence to support recommendations on limiting sedentary behaviours, but there is still uncertainty about the dose–response relationship as well as about variations by type or domain of sedentary behaviour [29]. To further understand how the different types of screen time, including nonsedentary screen time, affect somatic and mental health outcomes, it is important to continue measuring the screen time quantity as well as types of screen usage over time.
4.3.2. Physical Activity Assessment
To make the physical activity assessment more objective and further address physical activity behaviour patterns, device‐based measurements are promising. However, the data collection process should be optimised for future assessments (digitalisation and reminders) to minimise missing data. Additionally, the sampling of the subpopulation would benefit from targeted recruitment to ensure inclusion of children and adolescents with lower socioeconomic status, and thereby increase the generalisability of the study findings.
4.4. Strengths and Limitations
A strength of our analysis is the focus on the postpandemic status of physical activity and screen time behaviour, which has not been studied on a population level in Sweden before. Another benefit of the Generation Pep study is that it enables analyses of physical activity behaviour of different age categories by including children as well as adolescents. Furthermore, device‐measured physical activity, as well as more in‐depth analysis of screen time usage introduced in 2022, provides valuable additional information.
A limitation is the use of self‐reported data for assessing PA and screen time. As reported previously, the questions used are based on the HBSC questionnaire (HBSC study | Health Behaviour in School‐aged Children study), and the questions for assessing PA and screen time have not been validated for the survey. The decision not to choose a comprehensive multi‐item validated questionnaire was based on the intention to decrease participation burden. In this context, it is also relevant to note that the primary purpose of the Generation Pep survey is to assess trends in PA and screen time over time, which will likely not be impacted by this limitation. Regardless of whether questions are being validated or not, as self‐reported measures always come with the risk of social desirability bias, the most accurate and precise way to assess physical activity would have been to use device‐based measurement. However, this was not deemed feasible for the whole survey population. Another limiting factor of the survey was the categorical nature of some questions (e.g., screen time), which did not allow us to draw conclusions about the exact effect sizes. However, according to our results, an additional 7% of the study population had a screen time exceeding the current recommendations of 1–2 h after the pandemic compared to the prepandemic phase. On a population level, this can be considered a relevant behaviour change, which might have implications on public health [7]. Due to the observational design of the study, no causal relationship between the pandemic phases and movement behaviour can be proven. There might have been other unmeasured confounding factors, such as increased availability of screen media devices, more opportunities for digital engagement or the rising prevalence of overweight, that have influenced screen time and physical activity during this time. Further, the educational level and household income of the caregivers in our cohort are above the Swedish average, which might affect generalisability of the results. This pattern is, however, anticipated, as nonresponse bias has been reported to be more prevalent among individuals who report poorer health and lower income levels [30].
4.5. Conclusion
Our findings from the annual Generation Pep survey showed that screen time of Swedish children and adolescents was higher during and after the COVID‐19 pandemic compared with prepandemic years. To understand the impact of our findings and interpret the trend in a broader context, also considering changes in the digital environment, continuous detailed monitoring of screen time and its association with health outcomes should be prioritised. Results for physical activity levels were more inconsistent but seemed to be less affected by the pandemic phases. Our findings further indicate that future studies would benefit from adding device‐measured physical activity.
Author Contributions
Nina Kägi‐Braun: writing – original draft, formal analysis, methodology, investigation, writing – review and editing. Linnea Hedin: methodology, writing – original draft, investigation, writing – review and editing. Sophie Cassel: writing – review and editing, investigation. Anders Carlander: investigation, writing – review and editing. Malin J‐Son Höök: investigation, writing – review and editing. Marie Löf: conceptualization, writing – original draft, investigation, methodology, funding acquisition, writing – review and editing.
Ethics Statement
This study has been approved by the Swedish Ethical Review Authority (2018/375–31/5; 2018/1461–32; 2019–03798; 2020–03853; 2021–03931; 2022–03926‐02; and 2023–04417‐02).
Consent
The authors have nothing to report.
Conflicts of Interest
The authors declare no conflicts of interest.
Supporting information
Data S1.
Acknowledgements
The authors would like to thank all the participating children and adolescents, as well as their parents. We would also like to thank Jairo Migueles for processing accelerometry data and Anette Jansson, Pia Lindeskog, Anna Karin Lindroos, Gisela Nyberg and Anders Raustorp (reference group) for valuable contributions to the study questionnaire.
Funding: This work was supported by the Swiss National Science Foundation and Swedish Hjärt‐Lungfonden.
Data Availability Statement
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.
References
- 1. WHO , “WHO Guidelines on Physical Activity and Sedentary Behaviour,” 2020. [PubMed]
- 2. Poitras V. J., Gray C. E., Borghese M. M., et al., “Systematic Review of the Relationships Between Objectively Measured Physical Activity and Health Indicators in School‐Aged Children and Youth,” Applied Physiology, Nutrition, and Metabolism 41, no. 6 Suppl. 3 (2016): S197–S239, 10.1139/apnm-2015-0663. [DOI] [PubMed] [Google Scholar]
- 3. Carson V., Hunter S., Kuzik N., et al., “Systematic Review of Sedentary Behaviour and Health Indicators in School‐Aged Children and Youth: An Update,” Applied Physiology, Nutrition, and Metabolism 41, no. 6 Suppl. 3 (2016): S240–S265, 10.1139/apnm-2015-0630. [DOI] [PubMed] [Google Scholar]
- 4. Sanders T., Noetel M., Parker P., et al., “An Umbrella Review of the Benefits and Risks Associated With Youths’ Interactions With Electronic Screens,” Nature Human Behaviour 8, no. 1 (2024): 82–99, 10.1038/s41562-023-01712-8. [DOI] [PubMed] [Google Scholar]
- 5. Guthold R., Stevens G. A., Riley L. M., and Bull F. C., “Global Trends in Insufficient Physical Activity Among Adolescents: A Pooled Analysis of 298 Population‐Based Surveys With 1·6 Million Participants,” Lancet Child & Adolescent Health 4, no. 1 (2020): 23–35, 10.1016/s2352-4642(19)30323-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Neville R. D., Lakes K. D., Hopkins W. G., et al., “Global Changes in Child and Adolescent Physical Activity During the COVID‐19 Pandemic: A Systematic Review and Meta‐Analysis,” JAMA Pediatrics 176, no. 9 (2022): 886–894, 10.1001/jamapediatrics.2022.2313. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Nilsen S. A., Stormark K. M., Heradstveit O., and Breivik K., “Trends in Physical Health Complaints Among Adolescents From 2014‐2019: Considering Screen Time, Social Media Use, and Physical Activity,” SSM ‐ Population Health 22 (2023): 101394, 10.1016/j.ssmph.2023.101394. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. World Health Organization , “Guidelines on Physical Activity, Sedentary Behaviour and Sleep for Children Under 5 Years of Age,” 2019. [PubMed]
- 9. Kreitz‐Sandberg S., ““We Have Our Lessons in Teams”—Strategies Chosen in Swedish Schools During the COVID‐19 Pandemic and Consequences for Students in Upper Secondary Education,” Journal für International und Interkulturell Vergleichende Erziehungswissenschaft 28 (2022): 250–273. [Google Scholar]
- 10. Delisle Nyström C., Alexandrou C., Henström M., et al., “International Study of Movement Behaviors in the Early Years (SUNRISE): Results From SUNRISE Sweden's Pilot and COVID‐19 Study,” International Journal of Environmental Research and Public Health 17, no. 22 (2020): 8491. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Delisle Nyström C., Carlander A., Cassel S., et al., “Physical Activity and Screen Time in Swedish Children and Adolescents: The Generation Pep Study 2018–2021,” Acta Paediatrica 112, no. 3 (2023): 460–468, 10.1111/apa.16594. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Berggren S., Almquist‐Tangen G., Wolfbrandt O., and Roswall J., “Effects of the COVID‐19 Pandemic on the Physical Activity and Screen Time Habits of Children Aged 11–13 Years in Sweden,” Frontiers in Public Health 11 (2023): 1241938, 10.3389/fpubh.2023.1241938. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Helgadóttir B., Fröberg A., Kjellenberg K., Ekblom Ö., and Nyberg G., “COVID‐19 Induced Changes in Physical Activity Patterns, Screen Time and Sleep Among Swedish Adolescents ‐ a Cohort Study,” BMC Public Health 23, no. 1 (2023): 380, 10.1186/s12889-023-15282-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Folkhälsomyndigheten , “Föreskrifter om Upphävande av Folkhälsomyndighetens Föreskrifter Och Allmänna Råd (HSLF‐FS 2021:87) om Smittskyddsåtgärder Mot Sjukdomen Covid‐19 HSLF‐FS 2022:14,” 2022.Chefsjurist Pär Ödman, Socialstyrelsen.
- 15. Rosell M., Carlander A., Cassel S., et al., “Generation Pep Study: A Population‐Based Survey on Diet and Physical Activity in 12,000 Swedish Children and Adolescents,” Acta Paediatrica 110, no. 9 (2021): 2597–2606, 10.1111/apa.15850. [DOI] [PubMed] [Google Scholar]
- 16. Milther C., Winther L., Stahlhut M., et al., “Validation of an Accelerometer System for Measuring Physical Activity and Sedentary Behavior in Healthy Children and Adolescents,” European Journal of Pediatrics 182, no. 8 (2023): 3639–3647, 10.1007/s00431-023-05014-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Trott M., Driscoll R., Irlado E., and Pardhan S., “Changes and Correlates of Screen Time in Adults and Children During the COVID‐19 Pandemic: A Systematic Review and Meta‐Analysis,” eClinicalMedicine 48 (2022): 101452, 10.1016/j.eclinm.2022.101452. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Steene‐Johannessen J., Hansen B. H., Dalene K. E., et al., “Variations in Accelerometry Measured Physical Activity and Sedentary Time Across Europe – Harmonized Analyses of 47,497 Children and Adolescents,” International Journal of Behavioral Nutrition and Physical Activity 17, no. 1 (2020): 38, 10.1186/s12966-020-00930-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Jago R., Salway R., House D., et al., “Short and Medium‐Term Effects of the COVID‐19 Lockdowns on Child and Parent Accelerometer‐Measured Physical Activity and Sedentary Time: A Natural Experiment,” International Journal of Behavioral Nutrition and Physical Activity 20, no. 1 (2023): 42, 10.1186/s12966-023-01441-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Salway R., De Vocht F., Emm‐Collison L., et al., “Comparison of Children's Physical Activity Profiles Before and After COVID‐19 Lockdowns: A Latent Profile Analysis,” PLoS One 18, no. 11 (2023): e0289344, 10.1371/journal.pone.0289344. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Steene‐Johannessen J., Anderssen S. A., Kolle E., et al., “Temporal Trends in Physical Activity Levels Across More Than a Decade – A National Physical Activity Surveillance System Among Norwegian Children and Adolescents,” International Journal of Behavioral Nutrition and Physical Activity 18, no. 1 (2021): 55, 10.1186/s12966-021-01120-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Nyberg G., Kjellenberg K., Fröberg A., and Lindroos A. K., “A National Survey Showed Low Levels of Physical Activity in a Representative Sample of Swedish Adolescents,” Acta Paediatrica 109, no. 11 (2020): 2342–2353, 10.1111/apa.15251. [DOI] [PubMed] [Google Scholar]
- 23. Dohrn I. M., Tarp J., Steene‐Johannessen J., Vasankari T., and Hagströmer M., “Device‐Measured Physical Activity and Sedentary Time in the Nordic Countries: A Scoping Review of Population‐Based Studies,” Journal of Sport and Health Science 13 (2024): 650–660, 10.1016/j.jshs.2024.04.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Santos R. M. S., Mendes C. G., Sen Bressani G. Y., et al., “The Associations Between Screen Time and Mental Health in Adolescents: A Systematic Review,” BMC Psychology 11, no. 1 (2023): 127, 10.1186/s40359-023-01166-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Boniel‐Nissim M., Marino C., Galeotti T., et al., A Focus on Adolescent Social Media Use and Gaming in Europe, Central Asia and Canada: Health Behaviour in School‐Aged Children International Report From the 2021/2022 Survey (World Health Organization, 2024). [Google Scholar]
- 26. Li X., Vanderloo L. M., Keown‐Stoneman C. D. G., et al., “Screen Use and Mental Health Symptoms in Canadian Children and Youth During the COVID‐19 Pandemic,” JAMA Network Open 4, no. 12 (2021): e2140875, 10.1001/jamanetworkopen.2021.40875. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Ma L., Evans B., Kleppang A. L., and Hagquist C., “The Association Between Screen Time and Reported Depressive Symptoms Among Adolescents in Sweden,” Family Practice 38 (2021): 773–779, 10.1093/fampra/cmab029. [DOI] [PubMed] [Google Scholar]
- 28. Madigan S., McArthur B. A., Anhorn C., Eirich R., and Christakis D. A., “Associations Between Screen Use and Child Language Skills: A Systematic Review and Meta‐Analysis,” JAMA Pediatrics 174, no. 7 (2020): 665–675, 10.1001/jamapediatrics.2020.0327. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Chaput J.‐P., Willumsen J., Bull F., et al., “WHO Guidelines on Physical Activity and Sedentary Behaviour for Children and Adolescents Aged 5–17 Years: Summary of the Evidence,” International Journal of Behavioral Nutrition and Physical Activity 17, no. 1 (2020): 141, 10.1186/s12966-020-01037-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Bollinger C. R., Hirsch B. T., Hokayem C. M., et al., “Trouble in the Tails? What We Know About Earnings Nonresponse 30 Years After Lillard,” Journal of Political Economy 127, no. 5 (2019): 2143–2185. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data S1.
Data Availability Statement
The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.