Advertisement

Associations of Passive and Active Screen Time With Psychosomatic Complaints of Adolescents

      Introduction

      Increased screen time is a ubiquitous part of adolescent life and is adversely associated with their well-being. However, it remains unclear whether different types of screen time have equivalent associations, or if relationships are dose-dependent.

      Methods

      The data were from 2 nationally representative Health Behaviour in School-aged Children (2010, 2014) surveys across 44 European and North American countries. Psychosomatic health was assessed using 8 complaints and dichotomized as high or low. Discretionary time spent on passive (e.g., TV) and mentally active (e.g., electronic games, computer use) screen-based activities was categorized into 3 groups. Data were analyzed in 2021.

      Results

      The study included 414,489 adolescents (average age, 13.6 [SD=1.63] years; 51.1% girls). Multilevel modeling showed that psychosomatic complaints increased monotonically once all forms of screen time exceeded 2 hours/day. Adolescents reporting high (>4 hours/day) TV time, compared with those reporting low (≤2 hours/day), had higher odds of reporting psychosomatic complaints with 67% higher odds (OR=1.67, 95% CI=1.62, 1.72) in boys and 71% (OR=1.71, 95% CI=1.66, 1.75) in girls. High electronic game use was associated with psychosomatic complaints, with odds being 78% higher in boys (OR=1.78, 95% CI=1.73, 1.84) and 88% higher in girls (OR=1.88, 95% CI=1.82, 1.94). Similar associations were found between computer use and psychosomatic complaints.

      Conclusions

      Passive and mentally active screen time are adversely associated with psychosomatic complaints in a dose-dependent manner, with associations slightly stronger for active than passive screen time. This study supports limiting any type of screen time, either passive or active, to 2 hours/day to foster well-being.

      INTRODUCTION

      Engagement with electronic screens among adolescents has increased in recent years. The time American teens spent online doubled between 2006 and 2016; by 2016, teens aged 17–18 years were spending 6 hours/day online during their leisure time, whereas those aged 15–16 years were spending 5 hours/day.
      • Twenge J
      • Martin G
      • Spitzberg B.
      Trends in U.S. adolescents’ media use. 1976‒2016: the rise of digital media, the decline of TV, and the (near) demise of print.
      Daily visits to social media sites increased from 52% in 2008 to 82% in 2016 in individuals aged 15–16 years with comparable increases in younger teens. Given the increase in screen use among iGen teens (born in 1995–2012), concerns have been raised whether high screen time (ST) is associated with poorer physical/psychosocial health among adolescents. ST is deleteriously associated with adiposity, cardiovascular fitness, depressive symptoms, and quality of life,
      • Stiglic N
      • Viner RM.
      Effects of screentime on the health and well-being of children and adolescents: a systematic review of reviews.
      ,
      • Tremblay MS
      • LeBlanc AG
      • Kho ME
      • et al.
      Systematic review of sedentary behaviour and health indicators in school-aged children and youth.
      supporting current guidelines for limiting children's ST.
      • Tremblay MS
      • Carson V
      • Chaput JP
      • et al.
      Canadian 24-hour movement guidelines for children and youth: an integration of physical activity, sedentary behaviour, and sleep.
      Department of Health
      Australian 24-hour movement guidelines for children and young people (5–17 years) - an integration of physical activity, sedentary behaviour and sleep.
      • Bull FC
      • Al-Ansari SS
      • Biddle S
      • et al.
      World Health Organization 2020 guidelines on physical activity and sedentary behaviour.
      Emerging evidence suggests a curvilinear relationship between ST and mental well-being, such that low ST may have some beneficial effects over abstinence/high use.
      • Liu M
      • Wu L
      • Yao S.
      Dose–response association of screen time-based sedentary behaviour in children and adolescents and depression: a meta-analysis of observational studies.
      ,
      • Suchert V
      • Hanewinkel R
      • Isensee B.
      Sedentary behavior and indicators of mental health in school-aged children and adolescents: a systematic review.
      A study of U.S. adolescents found that modest engagement with digital media, compared with above-average use, was not associated with well-being, whereas excessive use had small negative associations.
      • Ferguson CJ.
      Everything in moderation: moderate use of screens unassociated with child behavior problems.
      Another study found an inverted “U”–shaped relationship between digital media engagement and psychosocial functioning of adolescents, suggesting that approximately 1–2 hours/day was just right for better functioning, whereas very low/very high levels were associated with poorer functioning.
      • Przybylski AK
      • Orben A
      • Weinstein N.
      Examining the relationship between digital screen engagement and psychosocial functioning in a confirmatory cohort study.
      Such evidence supports the Goldilocks hypothesis that screen use in moderation may be just right, but contradicts the current ST guidelines and challenges the less-is-better hypothesis as a preventive measure for mental well-being.
      Given the varied use of screens in education, social networks, and everyday experiences, a better understanding of health associations of various types of screen-based activities may be more insightful than simply focusing on total time spent on screens.
      • Ferguson CJ.
      Everything in moderation: moderate use of screens unassociated with child behavior problems.
      It is possible that mentally active ST (e.g., video games) may be more cognitively demanding and provide stimulation that may not be achieved through mentally passive ST (e.g., TV), which may be linked with adverse health outcomes.
      • Werneck AO
      • Hoare E
      • Stubbs B
      • van Sluijs EMF
      • Corder K.
      Association of mentally-active and mentally-passive sedentary behaviour with depressive symptoms among adolescents.
      Furthermore, although computer and social media use can increase opportunities for cyberbullying and upward social comparison, watching TV can also increase such opportunities because TV broadcasting still has several connections with social media.
      • Behm-Morawitz E
      • Lewallen J
      • Miller B.
      Real mean girls? Reality television viewing, social aggression, and gender-related beliefs among female emerging adults.
      It is therefore important to understand how youth are engaging with digital devices, on which platforms and in what context.
      • Belfort EL.
      Editorial: Did Goldilocks have it right? How do we define too little, too much, or just right?.
      Existing evidence has primarily emerged from adult populations,
      • Hallgren M
      • Dunstan DW
      • Owen N.
      Passive versus mentally active sedentary behaviors and depression.
      ,
      • Hallgren M
      • Vancampfort D
      • Owen N
      • et al.
      Prospective relationships of mentally passive sedentary behaviors with depression: mediation by sleep problems.
      and relationships between various types of ST and mental well-being during adolescence remain unclear.
      Light video game use in children and adolescents was associated with the fewer externalizing and internalizing symptoms, whereas high use was marginally associated with more symptoms.
      • Przybylski AK.
      Electronic gaming and psychosocial adjustment.
      In contrast, TV viewing was negatively associated with the development of physical and cognitive abilities, and positively associated with obesity, sleep problems, depression, and anxiety in children.
      • Domingues-Montanari S.
      Clinical and psychological effects of excessive screen time on children.
      A recent review indicated the impact of TV viewing on anxiety and depression may not be as strong as other types of ST; however, the results were largely mixed.
      • Zink J
      • Belcher BR
      • Imm K
      • Leventhal AM.
      The relationship between screen-based sedentary behaviors and symptoms of depression and anxiety in youth: a systematic review of moderating variables.
      Evidence suggests that passive ST was associated with mood and anxiety disorders, whereas mentally active ST was not.
      • Kim S
      • Favotto L
      • Halladay J
      • Wang L
      • Boyle MH
      • Georgiades K.
      Differential associations between passive and active forms of screen time and adolescent mood and anxiety disorders.
      Earlier research also showed that passive ST during adolescence was associated with elevated psychological distress during adulthood, but mentally active ST was not.
      • Werneck AO
      • Hoare E
      • Stubbs B
      • van Sluijs EMF
      • Corder K.
      Associations between mentally-passive and mentally-active sedentary behaviours during adolescence and psychological distress during adulthood.
      Although these studies offer differential associations for various types of ST, they are based on single-country, nonrepresentative samples that are lacking in rich covariate data.
      • Belfort EL.
      Editorial: Did Goldilocks have it right? How do we define too little, too much, or just right?.
      A more recent multi-country study reported associations between total ST and mental well-being indicators without considering types of ST.
      • Khan A
      • Lee EY
      • Rosenbaum S
      • Khan SR
      • Tremblay MS.
      Dose-dependent and joint associations between screen time, physical activity, and mental wellbeing in adolescents: an international observational study.
      Using representative samples from multiple countries, in this study we aimed to examine associations of passive and mentally active ST with mental health indicators in adolescents by gender and whether the relationships are dose-dependent. Though overall ST is similar across genders, gender-based heterogeneity has been observed in ST types; thus, implications to health might also be different by gender, justifying gender stratification.
      • Twenge JM
      • Farley E.
      Not all screen time is created equal: associations with mental health vary by activity and gender.
      ,
      • Bucksch J
      • Sigmundova D
      • Hamrik Z
      • et al.
      International trends in adolescent screen-time behaviors from 2002 to 2010.

      METHODS

      Study Sample

      The data were from the Health Behaviour in School-aged Children study, a repeated cross-sectional survey aimed at monitoring adolescent health and well-being in Europe and North America.
      • Roberts C
      • Freeman J
      • Samdal O
      • et al.
      The Health Behaviour in School-aged Children (HBSC) study: methodological developments and current tensions.
      Using stratified random cluster sampling, the school-based data are collected every 4 years from a nationally representative sample of adolescents aged 11, 13, and 15 years in participating countries. Participants provide self-reported data by anonymously completing a questionnaire that includes a range of items on health indicators and related behaviors.
      • Roberts C
      • Freeman J
      • Samdal O
      • et al.
      The Health Behaviour in School-aged Children (HBSC) study: methodological developments and current tensions.
      This study used 2 consecutive rounds of Health Behaviour in School-aged Children data collected in 2010 and 2014 from 44 countries. Of the 427,675 participants, 414,489 adolescents had complete outcome data for the analytical sample. Ethics approvals from appropriate regulatory bodies were received in each country, and informed consent was obtained from participants and a parent/guardian. Analyses for this manuscript were cleared by The University of Queensland Human Ethics Committee (2021/HE000671).

      Measures

      Psychosomatic health complaints were assessed using the frequency participants experienced the following 8 complaints over the past 6 months: headache, abdominal pain, backache, feeling low, irritability or bad mood, feeling nervous, sleeping difficulties, and dizziness, with 5 response options: about every day, more than once a week, about every week, about every month, and rarely or never. This scale has adequate test–retest reliability and validity.
      • Haugland S
      • Wold B.
      Subjective health complaints in adolescence–reliability and validity of survey methods.
      In this study, the 8 items had good internal consistency (Cronbach's α=0.82). The scale was dichotomized as high or low (1: ≥2 complaints more than once a week; 0: otherwise)
      • Walsh SD
      • Sela T
      • De Looze M
      • et al.
      Clusters of contemporary risk and their relationship to mental well-being among 15-year-old adolescents across 37 countries.
      to identify how various levels of screen use were associated with high psychosocial complaints and their possible dose-dependent relationships.
      Discretionary time spent on screen-based activities was assessed using 3 self-reported items for weekdays and weekend days. Passive ST was assessed using the item: About how many hours a day do you usually watch television (including DVDs and videos) in your free time? Mentally active ST was assessed using the items: About how many hours a day do you usually play games on a computer or games console (PlayStation, Xbox, GameCube, etc.) in your free time? and About how many hours a day do you usually use a computer for chatting online, Internet, e-mailing, homework, etc., in your free time? Responses included 9 options: none at all, 0.5 hours/day, 1 hour/day, 2 hours/day, 3 hours/day, 4 hours/day, 5 hours/day, 6 hours/day, and ≥7 hours/day. These items have acceptable test–retest reliability.
      • Bobakova D
      • Hamrik Z
      • Badura P
      • Sigmundova D
      • Nalecz H
      • Kalman M.
      Test–retest reliability of selected physical activity and sedentary behaviour HBSC items in the Czech Republic, Slovakia and Poland.
      Items on weekdays and weekend days were weighted 5:2 to generate an average ST/day. Average ST/day was collapsed into 3 categories (≤2 hours/day, >2–4 hours/day, and >4 hours/day)
      • Kim S
      • Favotto L
      • Halladay J
      • Wang L
      • Boyle MH
      • Georgiades K.
      Differential associations between passive and active forms of screen time and adolescent mood and anxiety disorders.
      and informed by their relationships with psychosomatic complaints (Figure 1).
      Figure 1
      Figure 1Associations of screen uses with psychosomatic complaints by gender, using generalized additive models.
      Note: Chi-square test of nonlinearity: p<0.001 for all types of screen uses. All models were adjusted for age, BMI z-scores, drunkenness, physical activity, FAS, and survey cycle.
      hrs/d, hours per day; FAS, Family Affluence Scale.
      Figure 1
      Figure 1Associations of screen uses with psychosomatic complaints by gender, using generalized additive models.
      Note: Chi-square test of nonlinearity: p<0.001 for all types of screen uses. All models were adjusted for age, BMI z-scores, drunkenness, physical activity, FAS, and survey cycle.
      hrs/d, hours per day; FAS, Family Affluence Scale.
      Covariates were selected based on availability and plausible connection to the outcome measure.
      • Khan A
      • Lee EY
      • Rosenbaum S
      • Khan SR
      • Tremblay MS.
      Dose-dependent and joint associations between screen time, physical activity, and mental wellbeing in adolescents: an international observational study.
      Sociodemographic covariates included age and lifetime drunkenness. Individual-level SES was measured with the Family Affluence Scale,
      • Currie C
      • Molcho M
      • Boyce W
      • Holstein B
      • Torsheim T
      • Richter M.
      Researching health inequalities in adolescents: the development of the Health Behaviour in School-Aged Children (HBSC) family affluence scale.
      which is a composite score based on items that assess the households’ number of cars and computers, bedroom sharing, and number of family holidays in the past year. BMI (kg/m2), converted into BMI z-scores using the WHO Child Growth Standards, was derived from self-reported height and weight. Participants reported the number of days in the past week that they participated in moderate-to-vigorous physical activities for ≥60 minutes.
      • Prochaska JJ
      • Sallis JF
      • Long B.
      A physical activity screening measure for use with adolescents in primary care.
      Missing values for the study factors and covariates ranged from 0.8% (age) to 17.9% (BMI) (Table 1). To minimize biases owing to the missing data, the authors implemented multiple imputations by chained equations. Twenty imputations were chosen based on the rule that the number should be at least as large as the percentage of missing data.
      • White IR
      • Royston P
      • Wood AM.
      Multiple imputation using chained equations: issues and guidance for practice.
      The imputed descriptive statistic values closely matched the observed values.
      Table 1Description of Study Sample From 44 Countries, HBSC Study, 2010‒2014 (N=414,489)
      CharacteristicsBoysGirls
      Total participants202,599211,890
      Mean age (SD)13.59 (1.63)13.58 (1.63)
      Age group: 11-year-olds57,869 (28.56)61,466 (29.01)
       13-year-olds70,150 (34.63)73,214 (34.55)
       15-year-olds72,763 (35.91)75,701 (35.73)
       Missing1,817 (0.90)1,509 (0.71)
      Been drunk: never148,009 (73.06)163,085 (76.97)
       Once19,811 (9.78)18,302 (8.64)
       Twice or more25,512 (12.59)21,624 (10.21)
       Missing9,267 (4.57)8,879 (4.19)
      Family affluence scale: Q147,442 (23.42)56,628 (26.73)
       Q267,439 (33.29)71,726 (33.85)
       Q333,692 (16.63)33,972 (16.03)
       Q441,429 (20.45)39,851 (18.81)
       Missing12,597 (6.22)9,713 (4.58)
      Mean BMI (SD)
      Percentage of missing values: BMI=17.9%.
      19.82 (3.57)19.38 (3.38)
      Mean TV time (SD)
      Percentage of missing values: TV=8.8%.
      hours/day
      2.53 (1.70)2.40 (1.62)
      Mean e-game time (SD)
      Percentage of missing values: e-game=10.2%.
      hours/day
      2.28 (1.87)1.34 (1.65)
      Mean computer time (SD)
      Percentage of missing values: Computer use=8.8%.
      hours/day
      2.05 (1.89)2.18 (1.91)
      Mean PA (SD)
      Percentage of missing values (≥60 minutes of PA)=1.9%. e-game, electronic game; HBSC, Health Behaviour in School-aged Children; PA, physical activity; Q, quartile.
      days/week
      4.42 (2.07)3.80 (2.03)
      Prevalence of high psychosomatic complaints (%)25.4239.46
      a Percentage of missing values: BMI=17.9%.
      b Percentage of missing values: TV=8.8%.
      c Percentage of missing values: e-game=10.2%.
      d Percentage of missing values: Computer use=8.8%.
      e Percentage of missing values (≥60 minutes of PA)=1.9%.e-game, electronic game; HBSC, Health Behaviour in School-aged Children; PA, physical activity; Q, quartile.

      Statistical Analysis

      To examine the pattern of associations between passive and mentally active ST and psychosomatic complaints without any parametric assumptions on the nature of the relationships, generalized additive models (GAMs) were used, adjusted for age, BMI z-scores, drunkenness, Family Affluence Scale, physical activity, and survey cycle. The GAM is a nonparametric model that allows nonlinear relationships to be modeled without specifying any functional form. Stata commands gam and gamplot were used to estimate the model parameters and to plot the estimated function. Interactions between ST items and gender were explored.
      To quantify the associations, the authors conducted multilevel logistic regression modeling that considered the nested structure of the data and had 3 levels (countries, schools, individual participants), using the runmlwin command via Stata, version 17SE. To avoid issues of multicollinearity among different ST types, each type of ST was modeled separately to understand their relationship with psychosomatic complaints, resulting in 3 models. The analyses were repeated for psychological (e.g., feeling low, irritable, nervousness, and sleeping difficulty) and somatic (e.g., headache, abdominal pain, backache, and dizziness) components to examine whether estimates were different across the 2 constructs. To examine whether different ST types offered differential estimates for sufficiently active (engaged in MVPA ≥60 minutes/day) versus insufficiently active adolescents, the analyses were repeated across physical activity status. All models were adjusted for the same covariates. Finally, sensitivity analyses were conducted using psychological complaints as scores to examine whether modeling continuous outcome impacted the results.

      RESULTS

      A comparison between those included in analyses (n=414,489, analytical sample) and those excluded (n=13,186) showed that the average age was similar across the 2 groups (13.6 vs 13.2 years), girls were over-represented (51.1% vs 40.0%), and average family affluence was marginally higher in the analytical sample (5.7 vs 5.1). Descriptive statistics of study participants are reported in Table 1. Overall prevalence of high psychosomatic complaints was 32.6%, with girls reporting more complaints than boys (39.5% vs 25.4%). Country-level descriptive statistics are presented in Appendix Table 1 (available online). Interactions between each type of ST and gender were significant (p<0.001), supporting the gender-stratified analyses.
      Figure 1 presents smoothed functions from GAMs of psychosomatic complaints as a function of ST/day. A visual examination of the plots indicated curvilinear associations between all ST types and psychosomatic complaints (chi-square tests of nonlinearity, p<0.001) for girls and boys. None of the associations were detrimental during the first hour. TV time had trivial associations with high psychosomatic complaints in the first 2 hours before the association became harmful for both girls and boys. Adverse associations of using computers started after the first hour for boys and girls. For playing electronic games, adverse associations started after 1 hour for girls and 1.5 hours for boys.
      Table 2 presents adjusted association estimates of different levels of ST with high psychosomatic complaints. Overall, odds of high psychosomatic complaints increased monotonically with an increase of each type of ST, with girls showing slightly higher estimates than boys. Odds of high complaints were 67% higher in boys (OR=1.67, 95% CI=1.62, 1.72) and 71% higher in girls (OR=1.71, 95% CI=1.66, 1.75) who had >4 hours/day of TV time, compared with ≤2 hours/day. Adolescents with >4 hours/day of playing electronic games had the highest odds of high complaints, with 78% higher odds in boys (OR=1.78, 95% CI=1.73, 1.84) and 88% higher odds in girls (OR=1.88, 95% CI=1.82, 1.94). The odds were 27% higher in boys (OR=1.27, 95% CI=1.24, 1.30) and 40% higher in girls (OR=1.40, 95% CI=1.36, 1.44) for >2–4 hours/day of electronic games. Compared with ≤2 hours/day of computer use, odds of high complaints were 84% higher in boys (OR=1.84, 95% CI=1.79, 1.90) and double in girls (OR=2.08, 95% CI=2.03, 2.14) who spent >4 hours/day on computers. The odds were 32% higher in boys (OR=1.32, 95% CI=1.28, 1.35) and 37% higher in girls (OR=1.37, 95% CI=1.34, 1.40) for using computers >2–4 hours/day. When all 3 ST measures were included in the same model, the resulting estimates demonstrated significant and dose-responsive relationships (Appendix Table 2, available online), similar to what have been shown by the original estimates when each ST measure was modeled separately; however, these new estimates were lower than the original estimates.
      Table 2Association Estimates of Screen Uses With High Psychosomatic Complaints of Adolescents by Gender
      Model #: screen typeHigh psychosomatic complaints
      BoysGirls
      Unadjusted,Adjusted,Unadjusted,Adjusted,
      crude OR (95% CI)AOR (95% CI)crude OR (95% CI)AOR (95% CI)
      Model 1
      Adjusted for age, BMI z-scores, drunkenness, physical activity, FAS, and survey cycle for each screen type.
      : watching television
       ≤2 hours/day
      Reference category. #, number; FAS, Family Affluence Scale, AOR, Adjusted odds ratio.
      1.001.001.001.00
       >2‒4 hours/day1.19 (1.16, 1.21)1.16 (1.13, 1.19)1.30 (1.27,1.32)1.21 (1.19, 1.24)
       >4 hours/day1.78 (1.73, 1.83)1.67 (1.62, 1.72)1.91 (1.86, 1.96)1.71 (1.66, 1.75)
      Model 2
      Adjusted for age, BMI z-scores, drunkenness, physical activity, FAS, and survey cycle for each screen type.
      : electronic games
       ≤2 hours/day
      Reference category. #, number; FAS, Family Affluence Scale, AOR, Adjusted odds ratio.
      1.001.001.001.00
       >2‒4 hours/day1.30 (1.27, 1.33)1.27 (1.24, 1.30)1.42 (1.39, 1.46)1.40 (1.36, 1.44)
       >4 hours/day1.89 (1.84, 1.94)1.78 (1.73, 1.84)2.07 (2.01, 2.14)1.88 (1.82, 1.94)
      Model 3
      Adjusted for age, BMI z-scores, drunkenness, physical activity, FAS, and survey cycle for each screen type.
      : computer uses
       ≤2 hours/day
      Reference category. #, number; FAS, Family Affluence Scale, AOR, Adjusted odds ratio.
      1.001.001.001.00
       >2‒4 hours/day1.35 (1.32, 1.39)1.32 (1.28, 1.35)1.55 (1.52, 1.59)1.37 (1.34, 1.40)
       >4 hours/day1.97 (1.91, 2.02)1.84 (1.79, 1.90)2.55 (2.49, 2.62)2.08 (2.03, 2.14)
      a Adjusted for age, BMI z-scores, drunkenness, physical activity, FAS, and survey cycle for each screen type.
      b Reference category.#, number; FAS, Family Affluence Scale, AOR, Adjusted odds ratio.
      The analyses showed that country and school levels, together, represented 6%–7% of total variability in the latent measurement of psychosomatic complaints. Country-level association estimates are presented in Appendix Table 3 (available online) with cross-country variations in the estimates. Compared with ≤2 hours/day, >4 hours/day of watching TV was associated with a significantly increase odds of high complains in 43/44 countries, >4 hours/day of electronic games was associated with a significantly increase odds of high complains in 42/43 countries, and >4 hours/day of computer use was associated with a significantly increase odds of high complains in 43/44 countries. Overall, association estimates for different types of ST were comparable across psychological and somatic constructs (Appendix Table 4, available online).
      Table 3Association Estimates of Screen Uses With High Psychosomatic Complaints by Physical Activity Status
      Model #: screen typeHigh psychosomatic complaints
      Sufficiently activeInsufficiently active
      Boys,Girls,Boys,Girls,
      AOR (95% CI)AOR (95% CI)AOR (95% CI)AOR (95% CI)
      Model 1
      Adjusted for age, BMI z-scores, drunkenness, physical activity, FAS, and survey cycle.
      : watching television
       ≤2 hours/day
      Reference category. #, number; FAS, Family Affluence Scale, AOR, Adjusted odds ratio.
      1.001.001.001.00
       >2‒4 hours/day1.21 (1.16, 1.27)1.29 (1.22, 1.36)1.14 (1.11, 1.17)1.20 (1.17, 1.22)
       >4 hours/day1.72 (1.63, 1.83)1.93 (1.80, 2.08)1.63 (1.58, 1.69)1.65 (1.61, 1.70)
      Model 2
      Adjusted for age, BMI z-scores, drunkenness, physical activity, FAS, and survey cycle.
      : electronic games
       ≤2 hours/day
      Reference category. #, number; FAS, Family Affluence Scale, AOR, Adjusted odds ratio.
      1.001.001.001.00
       >2‒4 hours/day1.40 (1.33, 1.47)1.44 (1.35, 1.55)1.23 (1.20, 1.27)1.39 (1.35, 1.43)
       >4 hours/day1.89 (1.78, 2.00)2.06 (1.89, 2.25)1.72 (1.66, 1.77)1.83 (1.76, 1.90)
      Model 3
      Adjusted for age, BMI z-scores, drunkenness, physical activity, FAS, and survey cycle.
      : computer uses
       ≤2 hours/day
      Reference category. #, number; FAS, Family Affluence Scale, AOR, Adjusted odds ratio.
      1.001.001.001.00
       >2‒4 hours/day1.39 (1.32, 1.46)1.42 (1.34, 1.51)1.29 (1.25, 1.33)1.36 (1.33, 1.40)
       >4 hours/day1.97 (1.86, 2.09)2.24 (2.09, 2.41)1.77 (1.71, 1.83)2.04 (1.98, 2.10)
      a Adjusted for age, BMI z-scores, drunkenness, physical activity, FAS, and survey cycle.
      b Reference category.#, number; FAS, Family Affluence Scale, AOR, Adjusted odds ratio.
      The association estimates for different ST types were slightly higher for sufficiently active than those for insufficiently active adolescents (Table 3). For example, watching TV >4 hours/day, compared with ≤2 hours/day, had 72% higher odds of high complaints in sufficiently active boys (OR=1.72, 95% CI=1.63, 1.83), whereas the odds were 63% higher in insufficiently active boys (OR=1.63, 95% CI=1.58, 1.69). Odds of high complaints were double in girls with sufficient physical activity (OR=2.06, 95% CI=1.89, 2.25) and 83% higher in girls with insufficient physical activity (OR=1.83, 95% CI=1.76, 1.90) for playing electronic games >4 hours/day.
      Sensitivity analyses with psychosomatic complaints as a continuous outcome produced similar results without meaningful changes (Appendix Table 5, available online). Adverse associations of watching TV, electronic games, and computer time with psychosomatic complaint scores were observed in a dose-dependent manner when compared with ≤2 hours/day as reference for girls and boys.

      DISCUSSION

      To the best of the authors’ knowledge, this large multi-country study is the first that provides evidence of dose-dependent adverse associations of passive and mentally active ST with psychosomatic complaints. Adverse associations were stronger for computer use and to some extent electronic games than they were for TV time, with estimates being slightly higher in girls than in boys. Potentially harmful effects of ST on psychosomatic complaints appear to begin about 30 minutes/day earlier with ST that requires mental exertion (i.e., electronic games, computer use) than passive use (i.e., TV time).
      The findings suggest that discretionary use of electronic screens is adversely associated with psychosomatic complaints at least after 1 hour/day for any screen use. This 1 hour/day timepoint was more apparent with playing electronic games and using computers than TV time, where the adverse association with psychosomatic complaints was observed at approximately 2 hours/day. Unlike previous studies asserting differential associations between varying types of ST and health outcomes,
      • Sanders T
      • Parker PD
      • Del Pozo-Cruz B
      • Noetel M
      • Lonsdale C.
      Type of screen time moderates effects on outcomes in 4013 children: evidence from the Longitudinal Study of Australian Children.
      ,
      • Stamatakis E
      • Coombs N
      • Jago R
      • et al.
      Type-specific screen time associations with cardiovascular risk markers in children.
      associations between different types of screens and psychosomatic complaints were largely nondifferential in this study. Nonetheless, a higher risk of psychosomatic complaints with higher exposure to screens builds on the growing evidence and requires further prospective investigations. Furthermore, although association estimates were nondifferential across the 3 screen types, smooth functions estimated by the GAMs indicated that different cut points could be considered in determining the critical window of intervention efforts. Specifically, mentally passive ST may not have a negative impact on psychosomatic complaints until approximately 2 hours/day of exposure whereas the psychosomatic tolerance level for ST that is more mentally involved may be shorter, showing an increase of psychosomatic complaints from approximately 1 hour/day of exposure. Country-level analyses showed detrimental associations of high ST (>4 hours/day) with psychosomatic complaints in almost all studied countries; however, moderate ST (>2–4 hours/day) had differential associations with psychosomatic complaints across countries. This needs to be considered when interpreting the study findings at the country-level.
      A gender difference in psychosomatic complaints was evident in this study, with girls reporting more complaints than boys (40% vs 25%). Earlier studies have indicated that girls are more likely to report poor mental well-being compared with boys,
      • Riecher-Rössler A.
      Sex and gender differences in mental disorders.
      • Schraedley PK
      • Gotlib IH
      • Hayward C.
      Gender differences in correlates of depressive symptoms in adolescents.
      • Van Voorhees BW
      • Paunesku D
      • Kuwabara SA
      • et al.
      Protective and vulnerability factors predicting new-onset depressive episode in a representative of U.S. adolescents.
      starting at puberty when mental health issues start to form, partly owing to endocrine and developmental changes within the complex sociocultural environment.
      • Sawyer SM
      • Afifi RA
      • Bearinger LH
      • et al.
      Adolescence: a foundation for future health.
      This study also showed nondifferential associations across gender; however, the strength of association was slightly higher in girls than in boys. Lack of gender differences may be because of the types of ST included in this study. Recent literature that includes social media/internet use suggests that these ST types are associated with poorer mental health among girls only.
      • Twenge JM
      • Farley E.
      Not all screen time is created equal: associations with mental health vary by activity and gender.
      Therefore, although girls typically report poorer mental well-being indicators than boys, the impact of ST may be similar across genders. Future prospective research should examine possible gender differences in such associations.
      The results of this study and cut points for safe discretionary ST exposure align well with some national guidelines,
      • Tremblay MS
      • Carson V
      • Chaput JP
      • et al.
      Canadian 24-hour movement guidelines for children and youth: an integration of physical activity, sedentary behaviour, and sleep.
      ,
      Department of Health
      Australian 24-hour movement guidelines for children and young people (5–17 years) - an integration of physical activity, sedentary behaviour and sleep.
      which recommend that school-aged children limit daily recreational ST to ≤2 hours/day. The WHO's most recent guidelines suggest limiting sedentary time throughout the day, particularly ST for recreational purposes.
      • Bull FC
      • Al-Ansari SS
      • Biddle S
      • et al.
      World Health Organization 2020 guidelines on physical activity and sedentary behaviour.
      The vague WHO recommendation stems largely from the lack of high-quality evidence on dose-responsiveness and sedentary behavior domain- or type-specific associations.
      • Chaput JP
      • Willumsen J
      • Bull F
      • et al.
      2020 WHO guidelines on physical activity and sedentary behaviour for children and adolescents aged 5-17 years: summary of the evidence.
      Findings of this study help fill this gap by investigating ST type-specific associations with psychosomatic complaints using nationally representative, multi-country, large samples of adolescents. However, further research is needed to be more conclusive on the ≤2 hours/day recreational ST cut point, with the exposure to mentally active ST potentially showing more susceptibility to psychosomatic complaints than passive ST.
      Strengths of this study include a large data set that includes representative samples from 44 European and North American countries, use of novel analytic techniques (e.g., GAMs), and adjustment for multiple covariates including physical activity. Analyses stratified by gender and sufficient/insufficient activity provide additional understanding of the associations. This study fills an important research gap on whether the association between ST and well-being is dose-dependent and varies by types of screen use, demonstrating a need for future consideration of screen use guidelines to consider ST types and their impact on varying health outcomes.

      Limitations

      Nevertheless, this study has limitations including the use of self-reported data, which are susceptible to biases, use of only psychosomatic complaints as the outcome, and lack of longitudinal individual-level or school-level contextual data that might explain the association. As adolescents may engage in various screen-based activities simultaneously, time spent on each screen type might have over-represented their actual use. Causality cannot be inferred based on cross-sectional data, and the possibility of reverse causation or bi-directional association cannot be ruled out.

      CONCLUSIONS

      Given that screen use is a ubiquitous part of iGen teens’ lives, clarifying the safe dose for different types of screen uses during the transition from childhood into young adulthood is an important challenge in sedentary behavior research. This study demonstrates adverse dose-dependent relationships between passive and mentally active ST and high psychosomatic complaints among adolescents of both genders. The adverse relationship of ST with psychosomatic complaints started when passive ST exceeded 1 hour/day and mentally active ST exceeded 2 hours/day. Though findings of this study largely support currently existing national guidelines
      • Tremblay MS
      • Carson V
      • Chaput JP
      • et al.
      Canadian 24-hour movement guidelines for children and youth: an integration of physical activity, sedentary behaviour, and sleep.
      ,
      • Bull FC
      • Al-Ansari SS
      • Biddle S
      • et al.
      World Health Organization 2020 guidelines on physical activity and sedentary behaviour.
      for limiting recreational ST to ≤2 hours/day, more prospective research is needed to confirm the findings of limiting each type of screen use at 2 hours/day to optimize mental well-being, which in turn can inform revision of the guidelines. Prolonged screen use may displace physical activity or sleep, which can be harmful for healthy development of adolescents.
      • Riecher-Rössler A.
      Sex and gender differences in mental disorders.
      Hence, a healthy balance of movement behaviors is desired when designing future public health strategies to optimize mental well-being of young people in high-income countries.

      ACKNOWLEDGMENTS

      Health Behaviour in School-aged Children (HBSC) is an international study carried out in collaboration with the WHO Regional Office for Europe. The HBSC Data Management Center is based at the Department of Health Promotion and Development in the University of Bergen, Norway. The authors thank the wider international HBSC network for developing the study, generating the data, and making them available for analyses. The corresponding author (Khan) ensures that the descriptions are accurate and agreed upon by all the authors.
      No financial disclosures were reported by the authors of this paper.

      CRediT AUTHOR STATEMENT

      Asaduzzaman Khan: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Software, Supervision, Validation, Visualization, Writing - original draft. Eun-Young Lee: Conceptualization, Investigation, Validation, Visualization, Writing - original draft. Ian Janssen: Methodology, Validation, Visualization, Writing - review and editing. Mark S. Tremblay: Investigation, Project administration, Supervision, Validation, Visualization, Writing - review and editing.

      Appendix. SUPPLEMENTAL MATERIAL

      REFERENCES

        • Twenge J
        • Martin G
        • Spitzberg B.
        Trends in U.S. adolescents’ media use. 1976‒2016: the rise of digital media, the decline of TV, and the (near) demise of print.
        Psychol Pop Media Cult. 2019; 8: 329-345https://doi.org/10.1037/ppm0000203
        • Stiglic N
        • Viner RM.
        Effects of screentime on the health and well-being of children and adolescents: a systematic review of reviews.
        BMJ Open. 2019; 9e023191https://doi.org/10.1136/bmjopen-2018-023191
        • Tremblay MS
        • LeBlanc AG
        • Kho ME
        • et al.
        Systematic review of sedentary behaviour and health indicators in school-aged children and youth.
        Int J Behav Nutr Phys Act. 2011; 8: 98https://doi.org/10.1186/1479-5868-8-98
        • Tremblay MS
        • Carson V
        • Chaput JP
        • et al.
        Canadian 24-hour movement guidelines for children and youth: an integration of physical activity, sedentary behaviour, and sleep.
        Appl Physiol Nutr Metab. 2016; 41 (suppl 3): S311-S327https://doi.org/10.1139/apnm-2016-0151
        • Department of Health
        Australian 24-hour movement guidelines for children and young people (5–17 years) - an integration of physical activity, sedentary behaviour and sleep.
        Department of Health, Australian Government, Canberra, Australia2021 (Published April. Accessed November 15, 2021)
        • Bull FC
        • Al-Ansari SS
        • Biddle S
        • et al.
        World Health Organization 2020 guidelines on physical activity and sedentary behaviour.
        Br J Sports Med. 2020; 54: 1451-1462https://doi.org/10.1136/bjsports-2020-102955
        • Liu M
        • Wu L
        • Yao S.
        Dose–response association of screen time-based sedentary behaviour in children and adolescents and depression: a meta-analysis of observational studies.
        Br J Sports Med. 2016; 50: 1252-1258https://doi.org/10.1136/bjsports-2015-095084
        • Suchert V
        • Hanewinkel R
        • Isensee B.
        Sedentary behavior and indicators of mental health in school-aged children and adolescents: a systematic review.
        Prev Med. 2015; 76: 48-57https://doi.org/10.1016/j.ypmed.2015.03.026
        • Ferguson CJ.
        Everything in moderation: moderate use of screens unassociated with child behavior problems.
        Psychiatr Q. 2017; 88: 797-805https://doi.org/10.1007/s11126-016-9486-3
        • Przybylski AK
        • Orben A
        • Weinstein N.
        Examining the relationship between digital screen engagement and psychosocial functioning in a confirmatory cohort study.
        J Am Acad Child Adolesc Psychiatry. 2020; 59: 1080-1088https://doi.org/10.1016/j.jaac.2019.06.017
        • Werneck AO
        • Hoare E
        • Stubbs B
        • van Sluijs EMF
        • Corder K.
        Association of mentally-active and mentally-passive sedentary behaviour with depressive symptoms among adolescents.
        J Affect Disord. 2021; 294: 143-150https://doi.org/10.1016/j.jad.2021.07.004
        • Behm-Morawitz E
        • Lewallen J
        • Miller B.
        Real mean girls? Reality television viewing, social aggression, and gender-related beliefs among female emerging adults.
        Psychol Pop Media Cult. 2016; 5: 340-355https://doi.org/10.1037/ppm0000074
        • Belfort EL.
        Editorial: Did Goldilocks have it right? How do we define too little, too much, or just right?.
        J Am Acad Child Adolesc Psychiatry. 2020; 59: 1025-1027https://doi.org/10.1016/j.jaac.2019.11.023
        • Hallgren M
        • Dunstan DW
        • Owen N.
        Passive versus mentally active sedentary behaviors and depression.
        Exerc Sport Sci Rev. 2020; 48: 20-27https://doi.org/10.1249/JES.0000000000000211
        • Hallgren M
        • Vancampfort D
        • Owen N
        • et al.
        Prospective relationships of mentally passive sedentary behaviors with depression: mediation by sleep problems.
        J Affect Disord. 2020; 265: 538-544https://doi.org/10.1016/j.jad.2019.11.088
        • Przybylski AK.
        Electronic gaming and psychosocial adjustment.
        Pediatrics. 2014; 134: e716-e722https://doi.org/10.1542/peds.2013-4021
        • Domingues-Montanari S.
        Clinical and psychological effects of excessive screen time on children.
        J Paediatr Child Health. 2017; 53: 333-338https://doi.org/10.1111/jpc.13462
        • Zink J
        • Belcher BR
        • Imm K
        • Leventhal AM.
        The relationship between screen-based sedentary behaviors and symptoms of depression and anxiety in youth: a systematic review of moderating variables.
        BMC Public Health. 2020; 20: 472https://doi.org/10.1186/s12889-020-08572-1
        • Kim S
        • Favotto L
        • Halladay J
        • Wang L
        • Boyle MH
        • Georgiades K.
        Differential associations between passive and active forms of screen time and adolescent mood and anxiety disorders.
        Soc Psychiatry Psychiatr Epidemiol. 2020; 55: 1469-1478https://doi.org/10.1007/s00127-020-01833-9
        • Werneck AO
        • Hoare E
        • Stubbs B
        • van Sluijs EMF
        • Corder K.
        Associations between mentally-passive and mentally-active sedentary behaviours during adolescence and psychological distress during adulthood.
        Prev Med. 2021; 145106436https://doi.org/10.1016/j.ypmed.2021.106436
        • Khan A
        • Lee EY
        • Rosenbaum S
        • Khan SR
        • Tremblay MS.
        Dose-dependent and joint associations between screen time, physical activity, and mental wellbeing in adolescents: an international observational study.
        Lancet Child Adolesc Health. 2021; 5: 729-738https://doi.org/10.1016/S2352-4642(21)00200-5
        • Twenge JM
        • Farley E.
        Not all screen time is created equal: associations with mental health vary by activity and gender.
        Soc Psychiatry Psychiatr Epidemiol. 2021; 56: 207-217https://doi.org/10.1007/s00127-020-01906-9
        • Bucksch J
        • Sigmundova D
        • Hamrik Z
        • et al.
        International trends in adolescent screen-time behaviors from 2002 to 2010.
        J Adolesc Health. 2016; 58: 417-425https://doi.org/10.1016/j.jadohealth.2015.11.014
        • Roberts C
        • Freeman J
        • Samdal O
        • et al.
        The Health Behaviour in School-aged Children (HBSC) study: methodological developments and current tensions.
        Int J Public Health. 2009; 54: 140-150https://doi.org/10.1007/s00038-009-5405-9
        • Haugland S
        • Wold B.
        Subjective health complaints in adolescence–reliability and validity of survey methods.
        J Adolesc. 2001; 24: 611-624https://doi.org/10.1006/jado.2000.0393
        • Walsh SD
        • Sela T
        • De Looze M
        • et al.
        Clusters of contemporary risk and their relationship to mental well-being among 15-year-old adolescents across 37 countries.
        J Adolesc Health. 2020; 66: S40-S49https://doi.org/10.1016/j.jadohealth.2020.02.012
        • Bobakova D
        • Hamrik Z
        • Badura P
        • Sigmundova D
        • Nalecz H
        • Kalman M.
        Test–retest reliability of selected physical activity and sedentary behaviour HBSC items in the Czech Republic, Slovakia and Poland.
        Int J Public Health. 2015; 60: 59-67https://doi.org/10.1007/s00038-014-0628-9
        • Currie C
        • Molcho M
        • Boyce W
        • Holstein B
        • Torsheim T
        • Richter M.
        Researching health inequalities in adolescents: the development of the Health Behaviour in School-Aged Children (HBSC) family affluence scale.
        Soc Sci Med. 2008; 66: 1429-1436https://doi.org/10.1016/j.socscimed.2007.11.024
        • Prochaska JJ
        • Sallis JF
        • Long B.
        A physical activity screening measure for use with adolescents in primary care.
        Arch Pediatr Adolesc Med. 2001; 155: 554-559https://doi.org/10.1001/archpedi.155.5.554
        • White IR
        • Royston P
        • Wood AM.
        Multiple imputation using chained equations: issues and guidance for practice.
        Stat Med. 2011; 30: 377-399https://doi.org/10.1002/sim.4067
        • Sanders T
        • Parker PD
        • Del Pozo-Cruz B
        • Noetel M
        • Lonsdale C.
        Type of screen time moderates effects on outcomes in 4013 children: evidence from the Longitudinal Study of Australian Children.
        Int J Behav Nutr Phys Act. 2019; 16: 117https://doi.org/10.1186/s12966-019-0881-7
        • Stamatakis E
        • Coombs N
        • Jago R
        • et al.
        Type-specific screen time associations with cardiovascular risk markers in children.
        Am J Prev Med. 2013; 44: 481-488https://doi.org/10.1016/j.amepre.2013.01.020
        • Riecher-Rössler A.
        Sex and gender differences in mental disorders.
        Lancet Psychiatry. 2017; 4: 8-9https://doi.org/10.1016/S2215-0366(16)30348-0
        • Schraedley PK
        • Gotlib IH
        • Hayward C.
        Gender differences in correlates of depressive symptoms in adolescents.
        J Adolesc Health. 1999; 25: 98-108https://doi.org/10.1016/S1054-139X(99)00038-5
        • Van Voorhees BW
        • Paunesku D
        • Kuwabara SA
        • et al.
        Protective and vulnerability factors predicting new-onset depressive episode in a representative of U.S. adolescents.
        J Adolesc Health. 2008; 42: 605-616https://doi.org/10.1016/j.jadohealth.2007.11.135
        • Sawyer SM
        • Afifi RA
        • Bearinger LH
        • et al.
        Adolescence: a foundation for future health.
        Lancet. 2012; 379: 1630-1640https://doi.org/10.1016/S0140-6736(12)60072-5
        • Chaput JP
        • Willumsen J
        • Bull F
        • et al.
        2020 WHO guidelines on physical activity and sedentary behaviour for children and adolescents aged 5-17 years: summary of the evidence.
        Int J Behav Nutr Phys Act. 2020; 17: 141https://doi.org/10.1186/s12966-020-01037-z