- Email: [email protected]
Screen-Based Behaviors of Children and Cardiovascular Risk Factors
Screen-Based Behaviors of Children and Cardiovascular Risk Factors Sarah Robinson, PhD, Robin M. Daly, PhD, Nicola D. Ridgers, PhD, and Jo Salmon, PhD Objective To determine whether the amount of time spent in screen-based behaviors (SBBs; television viewing, computer use, and playing electronic games) is independently associated with individual and clustered cardiovascular disease (CVD) risk factors among elementary school children. Study design Baseline data were used from 264 children (age 7-10 years) participating in the Transform-Us! cluster-randomized controlled trial. Time (h/d) spent in SBBs was obtained using a parent proxy-report questionnaire. Anthropometrics, blood pressure (BP), and lipids were measured using standard techniques. A clustered CVD risk score was calculated as the average of the standardized values of the subcomponents (waist circumference [WC], systolic BP, diastolic BP, and lipids). Results After adjusting for sex, parent education, physical activity (accelerometry), diet, and WC (when adiposity was not the outcome), television viewing time was positively associated with body mass index z-score (P = .002), WC (P = .02), and systolic BP (P = .05). Electronic games was positively associated with low density lipoprotein levels (P = .05), and total screen-time was positively associated with body mass index (P = .02). Conclusions Differential associations were observed between types of SBBs and CVD risk factors, indicating that not all SBBs are adversely associated with obesity and CVD risk. There is a need to differentiate between types of SBBs when evaluating the CVD risk associated with screen behaviors in children. (J Pediatr 2015;-:---). Trial registration International Standard Randomized Controlled Trial: ISRCTN83725066; Australian New Zealand Clinical Trials Registry: ACTRN12609000715279.
S
creen-based behaviors (SBBs; television [TV] viewing, computer use, and playing electronic games [e-games]) are key children’s sedentary leisure time pursuits. The American Academy of Pediatrics and government health authorities in the US, Australia, and Canada recommends limiting recreational screen-time to no more than 2 hours per day.1-4 However, population estimates in these countries show that between 40% and 80% of young people are exceeding screen-time recommendations.5-7 Since the initial establishment of pediatric screen-time recommendations in the US,2 which were based on time spent watching TV, pursuits such as using a computer and playing sedentary e-games have become highly pervasive among young people. Sedentary e-games and computer use may have physiological and metabolic effects that are not comparable with the more passive nature of watching TV and may have different behavioral mediators, such as snacking, which has been associated with cardiovascular disease (CVD) risk factors.8 To date, TV viewing and obesity are the dominant measures used in studies of screentime and CVD risk factors. A limited number of studies have examined the association between e-games and computer use with blood pressure (BP)8-10 and lipids9,11 in primary school children, and the findings are inconclusive. Few studies have accounted for the potential confounding effects of moderate-to-vigorous physical activity (MVPA) and diet (energy) quality, despite being independent predictors of CVD risk factors.9 Consequently, comparability between studies is restricted and the true CVD risk associated with different SBBs remains uncertain. Therefore, the aim of this study was to examine the independent associations between distinct SBBs (TV viewing, e-games, and computer use) and CVD risk factors and a clustered CVD risk score in a cohort of 7- to 10-year-old Australian children.
BMI BP CVD DBP e-games EYHS HDL-C LDL-C
Body mass index Blood pressure Cardiovascular disease Diastolic blood pressure Electronic games European Youth Heart Study High density lipoprotein cholesterol Low density lipoprotein cholesterol
MVPA NHANES SBB SBP SES TC TG TV WC
Moderate-to-vigorous physical activity National Health and Nutrition Examination Survey Screen-based behavior Systolic blood pressure Socioeconomic status Total cholesterol Triglycerides Television Waist circumference
From the Center for Physical Activity and Nutrition Research, School of Exercise and Nutrition Sciences, Deakin University, Burwood, Victoria, Australia Transform-Us! was supported by the National Health and Medical Research Council of Australia (2009-2013; ID533815) and the Diabetes Australia Research Trust. S.R. is supported by the National Heart Foundation (postgraduate scholarship). N.R. is supported by the Australian Research Council (Discovery Early Career Researcher Award DE120101173). J.S. is supported by the National Health and Medical Research Council (Principal Research Fellowship APP1026216). The authors declare no conflicts of interest. 0022-3476/$ - see front matter. Copyright ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2015.08.067
1
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Methods Data were drawn from baseline of the Transform-Us! (ACTRN12609000715279; ISRCTN83725066) intervention trial, which has been described in detail elsewhere.12 In brief, Transform-Us! was a cluster-randomized controlled trial that aimed to reduce sedentary behavior, increase physical activity, and improve children’s health. Twenty primary (elementary) schools within a 50 km (31 miles) radius of the city of Melbourne, Australia, with an enrolment of over 300 students stratified by socioeconomic status (SES; 8 of 41 schools in low SES areas and 12 of 96 schools in mid-high SES areas) were randomly selected. All children in year 3 (n = 1606) at the participating schools were invited to take part. Fivehundred ninety-one children (37%) and 446 parents (28%) completed baseline assessments, which were used for the present study. Information was not obtained concerning nonresponders; it is an ethics requirement in Australia for active informed consent to be provided. A subsample of 344 children (21%) consented to provide a blood sample, and 219 of those children completed the assessment. Two-hundred sixty-four children had at least one CVD risk factor measure. Ethics approval was obtained from the Deakin University Human Research Ethics Committee, the Victorian Department of Education and Early Childhood Development, and the Catholic Education Office Melbourne. All parents were informed of the study purpose prior to randomization, therefore, any changes in children’s SBBs as a consequence of this information would have been consistent across the whole sample enabling the data to be pooled at baseline for analysis. The time spent in 3 types of sedentary SBBs: (1) TV viewing, videos, and DVDs; (2) nonactive e-games (eg, PlayStation, Nintendo) and computer games; and (3) computer and internet use (excluding games) were reported by the main parent or caregiver (referred to here-on in as the parent) of the child. Data concerning whether or not the child usually participated in a range of leisure activities during a typical week in the current school term (yes/no), the total hours and minutes from Monday to Friday, and the total hours and minutes on Saturday and Sunday were collected.13 Total screen-time (h/d) was calculated as the sum of time spent in TV viewing, e-games, and computer use. Height, weight, and waist circumference (WC) were measured by trained research staff at the child’s school. Height was measured without shoes to the nearest 0.1 cm using a portable stadiometer (Seca 220; Seca, Hamburg, Germany). Weight was measured in light clothing to the nearest 0.1 kg using portable electronic Tanita scales (Wedderburn 1582; Wedderburn, Sydney, Australia). Each was measured twice, and the average of the two measures determined. Body mass index (BMI) was calculated from the average of the height and weight measures as kg/m2 to classify overweight ($25 kg/ m2) and obesity ($30 kg/m2) according to International Obesity Task Force age- and sex-specific cut-points that correspond to BMI values in adults.14 WC was measured using a flexible steel tape at the narrowest point between the 2
Vol. -, No. bottom rib and the iliac crest, in the midaxillary plane over light clothing.15 Age- and sex-specific percentiles were derived using Australian specific data.16 A WC $90th percentile was used to classify overweight and obesity.17 Resting BP was measured in accordance with standard procedures and recommendations.18 After 2 minutes of seated quiet rest, BP was measured on the right arm with an automatic digital BP machine (Omron Hem-907; Omron, Melbourne, Australia) using an appropriately sized pediatric cuff. Three measurements were taken at 1-minute intervals on two occasions, one week apart (total of 6 readings). The first measurement was discarded from each visit, and the average was calculated from the remaining 4 measurements. Systolic (SBP) and diastolic BP (DBP) percentiles were determined using height percentiles for sex and age.19 Elevated BP was defined as SBP and/or DBP $90th and <95th percentile; hypertension was defined as SBP and/or DBP $95th percentile.18 Children were provided with Eutectic Mixture of Local Anaesthetics anesthetic cream and instructed to attend a local pathology clinic to provide a fasting morning blood sample. Total cholesterol (TC), high density lipoprotein cholesterol (HDL-C), low density lipoprotein cholesterol (LDL-C), and triglycerides (TG) were assessed using standard techniques at a National Association of Testing Authorities/Royal College of Pathologists Australasia accredited pathology laboratory. Ageand sex-specific percentile cut-points were used to determine children at risk: elevated TC, LDL-C, and TG were defined as $90th percentile and low HDL-C as #10th percentile.20 Based on previous cardiometabolic risk score methodology from the European Youth Heart Study (EYHS),9 a continuous clustered CVD risk score was derived using the following 6 variables: SBP, DBP, LDL-C, HDL-C, TG, and WC. The calculation for average SBP and average DBP are described above, and these variables were used in the calculation of the CVD risk score. As HDL-C is inversely related to cardiovascular risk it was multiplied by 1 to indicate higher CVD risk with increasing value. For each of the 6 variables, the z-score was calculated as the number of SD units from the sample mean after normalization of the variable (Z = [value mean]/SD). All standardized scores were summed to create a clustered CVD risk score with a higher score indicating a higher level of risk. An estimate of children’s consumption of 6 key energy dense food types (salty snacks, chocolate and sweets, cakes, pastries, fast food, chips) and two beverage types (fruit juice and soft drinks) was obtained from a parental-proxy report. These 8 items have previously been identified as important contributors to children’s energy intake.21 Responses for food items were assessed on a monthly 9-point scale ranging from “never or less than once a month” (score = 1) to “6 or more times a day” (score = 9). Responses for beverage items were assessed on a daily 8-point scale ranging from “my child does not drink this beverage” (score = 1) to “6 or more servings a day” (score = 8). The frequency for each item was summed to provide a total diet energy density score with a higher score indicating greater energy density of the diet.21 The responding parent was also asked to select their highest Robinson et al
- 2015
ORIGINAL ARTICLES
level of education (never attended school to university or tertiary qualification). Children wore an ActiGraph GT3X accelerometer (ActiGraph, Pensacola, Florida) for 8 consecutive days on a belt positioned on the right hip during waking hours, except during water-based activities. The normal frequency filter was selected, and the epoch length was 15 seconds. MVPA (min/d) was defined using age-adjusted cut-points22 defined as $4 METs.23 Nonwear time is commonly defined in the literature as 20 minutes of consecutive zeros, and this definition was used in the current study.24 Wear time criterion was defined as 3 valid weekdays ($8 hours) and 1 valid weekend day ($7 hours).25 Data Analyses All analyses were undertaken using Stata/SE v 12.0 (StataCorp, College Station, Texas). The distributions of each of the outcome variables were assessed for normality. Because BMI was positively skewed, age- and sex-adjusted BMI zscores14 were calculated and used in all analyses. Descriptive data (mean, SD) were calculated for CVD risk factors. There were no differences in the anthropometric and SBP or DBP measures between children who provided a blood sample and those who did not (data not shown); therefore, all children with these health data were included in the analyses. Forced entry linear regression analyses were used to examine the association between daily TV viewing, e-games, computer use, and total screen-time (h/d) and CVD risk factors. Four statistical models were used to explore the associations and adjust for important confounding factors: model 1 adjusted for sex, school clustering, and parent education; model 2 additionally adjusted for MVPA and accelerometer wear time; model 3 additionally adjusted for diet energy density; and model 4 additionally adjusted for WC (where adiposity was not the outcome of interest). Only children with data for all covariates were included in analyses (Table I).
Results Characteristics and the number of children (enrolled February to June 2010) providing complete data for each variable are shown in Table I. The mean age was 8.7 (0.4) years, and approximately one-half of the parents had completed a university degree. Nearly 60% of children exceeded screentime recommendations of >2 hours per day. Approximately 1 in 5 children were classified as overweight or obese based on BMI (Table II). A high percentage of children had elevated TC levels (37.5%); however, only 2 children had HDL-C below the 10th percentile. Seventeen percent of the children had a SBP or DBP greater than the 90th percentile. TV Viewing TV viewing time was positively associated with BMI z-score (P = .002) and WC (P = .02) after adjusting for sex of the child, parent education, MVPA, accelerometer wear time, and diet
Table I. Participant characteristics Measures
Valid (n)
Mean (SD)
Age (y) Parent education (% university) Accelerometer MVPA (min/d) Accelerometer wear time (min/d) SBBs TV (min/d) E-games (min/d) Computer (min/d) Total screen (min/d) Screen-time >2 h/d (%) Diet Diet composite score Anthropometry Height (cm) Weight (kg) BMI (kg/m2) BMI z-score WC (cm) BP SBP (mm Hg) DBP (mm Hg) Cholesterol and lipids TC (mmol/L) HDL-C (mmol/L) LDL-C (mmol/L) TG (mmol/L) Clustered risk score*
264 264
8.7 (0.4) 46.2
264 264
66.9 (19.5) 699.0 (53.5)
262 264 263 261 261
101.5 (56.1) 25.4 (32.9) 19.6 (21.8) 146.4 (78.9) 58.6
264
21.0 (5.1)
263 263 263 263 264
132.0 (6.5) 30.0 (6.1) 17.1 (2.5) 0.2 (0.9) 59.1 (6.3)
255 255
101.7 (9.2) 60.1 (8.02)
147 147 147 147 147
4.59 (0.78) 1.69 (0.33) 2.58 (0.73) 0.69 (0.25) 0.29 (2.7)
All values represent mean (SD) or proportion (percentage). *Standardized and sum of z-scores for WC, the average of SBP and DBP, LDL-C, inverted HDLC, and TG.
energy density (Table III). A significant positive association was also found between TV viewing time and SBP (P = .05) after adjusting for sex, parent education, MVPA, accelerometer wear time, diet energy density, and WC. E-Games Time spent playing e-games was positively associated with LDL-C (P = .05) in the fully adjusted model. There was no association between e-games and any of the other cholesterol and lipid measures, BP, adiposity, or clustered CVD risk in any of the models. Computer Use Time spent using a computer was not associated with any of the CVD risk factors or clustered CVD risk scores in any of the models. Table II. Percentage of children with elevated levels of CVD risk factors CVD risk factors 2
BMI (>25 kg/m ) WC $75th percentile WC $90th percentile DBP and/or SBP >90th to <95th percentile DBP and/or SBP $95th percentile TC >90th percentile HDL-C <10th percentile LDL-C >90th percentile TG $90th percentile
Screen-Based Behaviors of Children and Cardiovascular Risk Factors
Valid (n)
%
263 264 264 255 255 147 147 147 147
22.1 30.7 12.9 6.6 10.5 37.5 1.4 12.9 19.7 3
THE JOURNAL OF PEDIATRICS
Vol. -, No. -
www.jpeds.com
Table III. Unstandardized regression coefficients of time (h/d) spent in SBBs and CVD risk factors TV
Adiposity (n = 261) z-BMI (kg/m2) Model 1 Model 2 Model 3 WC (cm) Model 1 Model 2 Model 3 BP (n = 255) SBP (mm Hg) Model 1 Model 2 Model 3 Model 4 DBP (mm Hg) Model 1 Model 2 Model 3 Model 4 Cholesterol and lipids (n = 146) HDL-C (mmol/L) Model 1 Model 2 Model 3 Model 4 LDL-C (mmol/L) Model 1 Model 2 Model 3 Model 4 TC (mmol/L) Model 1 Model 2 Model 3 Model 4 TG (mmol/L) Model 1 Model 2 Model 3 Model 4 CVD risk score* Model 1 Model 2 Model 3
E-games b (95% CI)
Computer P
b (95% CI)
Total screen
b (95% CI)
P
P
b (95% CI)
P
0.22 (0.08, 0.34) 0.22 (0.09, 0.36) 0.25 (0.10, 0.39)
.004 .003 .002
0.00 ( 0.29, 0.28) 0.01 ( 0.28, 0.26) 0.00 ( 0.29, 0.28)
.97 .92 .99
0.18 ( 0.50, 0.14) 0.18 ( 0.49, 0.14) 0.17 ( 0.49, 0.14)
.26 .26 .26
0.10 (0.01, 0.19) 0.10 (0.01, 0.19) 0.12 (0.02, 0.21)
.04 .04 .02
1.18 (0.09, 2.26) 1.24 (0.14, 2.34) 1.42 (0.29, 2.55)
.04 .03 .02
0.05 ( 2.08, 1.98) 0.16 ( 2.06, 1.73) 0.01 ( 2.12, 2.15)
.96 .86 .99
0.74 ( 3.08, 1.61) 0.67 ( 2.96, 1.62) 0.61 ( 2.84, 1.61)
.52 .55 .57
0.57 ( 0.20, 1.33) 0.58 ( 0.16, 1.33) 0.71 ( 0.08, 1.51)
.14 .12 .07
1.82 (0.23, 3.40) 1.84 (0.23, 3.45) 1.83 (0.18, 3.49) 1.52 (0.00, 3.04)
.03 .03 .03 .05
1.22 ( 1.32 ( 1.71 ( 1.74 (
3.72, 1.27) 3.86, 1.20) 4.14, 0.72) 4.04, 0.56)
.32 .29 .16 .13
1.76 ( 1.79 ( 1.69 ( 1.88 (
2.10, 5.63) 2.17, 5.74) 2.16, 5.55) 1.90, 5.67)
.35 .36 .37 .31
0.86 ( 0.85 ( 0.81 ( 0.60 (
0.36, 2.07) 0.37, 2.07) 0.41, 2.04) 0.58, 1.79)
.16 .16 .18 .30
0.54 ( 0.63 ( 0.60 ( 0.38 (
0.78, 1.87) 0.77, 2.03) 0.92, 2.11) 1.14, 1.91)
.40 .36 .42 .60
0.55 ( 0.89 ( 1.08 ( 1.09 (
3.18, 2.07) 3.32, 1.54) 3.59, 1.44) 3.66, 1.47)
.66 .45 .38 .39
0.80 ( 1.95, 3.54) 0.91 ( 1.90, 3.72) 0.87 (1.96, 3.71) 0.99 ( 1.81, 3.78)
.55 .51 .53 .47
0.28 ( 0.27 ( 0.23 ( 0.10 (
0.76, 1.31) 0.79, 1.33) 0.91, 1.38) 1.07, 1.28)
.59 .60 .68 .86
0.00 ( 0.01 ( 0.01 ( 0.03 (
0.06, 0.06) 0.06, 0.06) 0.06, 0.05) 0.02, 0.07)
.96 .86 .83 .31
0.06 ( 0.04 ( 0.04 ( 0.03 (
0.14, 0.01) 0.11, 0.02) 0.12, 0.03) 0.10, 0.04)
.10 .15 .19 .35
0.10 ( 0.09 ( 0.09 ( 0.06 (
0.08, 0.28) 0.12, 0.29) 0.11, 0.29) 0.13, 0.24)
.25 .38 .38 .52
0.00 ( 0.00 ( 0.00 ( 0.01 (
0.39, 0.04) 0.04, 0.04) 0.04, 0.04) 0.02, 0.05)
.97 .96 .93 .42
0.01 ( 0.02 ( 0.03 ( 0.00 (
0.12, 0.13) 0.10, 0.13) 0.08, 0.14) 0.10, 0.10)
.92 .76 .57 .99
0.12 ( 0.02, 0.27) 0.09 ( 0.04, 0.24) 0.13 (0.01, 0.25) 0.11 (0.00, 0.22)
.09 .16 .04 .05
0.04 ( 0.01 ( 0.00 ( 0.04 (
0.35, 0.28) 0.32, 0.30) 0.32, 0.33) 0.28, 0.35)
.81 .94 .98 .81
0.02 ( 0.02 ( 0.04 ( 0.02 (
0.06, 0.10) 0.05, 0.10) 0.04, 0.11) 0.05, 0.09)
.57 .49 .29 .49
0.02 ( 0.02 ( 0.04 ( 0.05 (
0.09, 0.12) 0.08, 0.13) 0.05, 0.13) 0.04, 0.13)
.72 .62 .39 .26
0.05 ( 0.03 ( 0.05 ( 0.04 (
0.10, 0.20) 0.12, 0.17) 0.09, 0.18) 0.11, 0.18)
.52 .72 .47 .58
0.04 ( 0.07 ( 0.09 ( 0.09 (
0.31, 0.40) 0.27, 0.41) 0.26, 0.43) 0.27, 0.44)
.80 .68 .61 .62
0.03 ( 0.03 ( 0.04 ( 0.05 (
0.05, 0.10) 0.04, 0.10) 0.02, 0.11) 0.01, 0.11)
.44 .35 .19 .14
0.01 ( 0.00 ( 0.00 ( 0.01 (
0.06, 0.04) 0.05, 0.04) 0.04, 0.04) 0.06, 0.03)
.61 .81 .99 .58
0.05 ( 0.04 ( 0.05 ( 0.04 (
0.03, 0.14) 0.06, 0.13) 0.05, 0.14) 0.05, 0.13)
.17 .43 .31 .37
0.07 ( 0.06 ( 0.05 ( 0.04 (
0.18, 0.04) 0.18, 0.07) 0.17, 0.07) 0.16, 0.07)
.19 .34 .38 .46
0.00 ( 0.00 ( 0.00 ( 0.00 (
0.04, 0.03) 0.03, 0.03) 0.03, 0.03) 0.31, 0.03)
.84 .91 .83 .84
0.28 ( 0.35, 0.91) 0.37 ( 0.23, 0.96) 0.46 ( 0.17, 1.10)
.36 .21 .14
0.61 ( 0.24, 1.45) 0.38 ( 0.36, 1.12) 0.55 ( 0.18, 1.29)
.15 .30 .13
0.86 ( 2.24, 0.51) 0.66 ( 2.14, 0.81) 0.58 ( 2.06, 0.89)
.20 .36 .42
0.17 ( 0.25, 0.59) 0.19 ( 0.21, 0.59) 0.28 ( 0.16, 0.71)
.41 .33 .19
b, beta coefficient. Model 1 adjusted for school clustering, SES, and sex of the child: model 2, additionally adjusted for MVPA and accelerometer wear time; model 3, additionally adjusted for a diet composite score; model 4, additionally adjusted for WC. Bold type indicates significance (P < .05). *Standardized and sum of WC, the average of SBP and DBP, LDL-C, inverted HDL-C, and TG.
Total Screen-Time Average daily total screen-time was positively associated with BMI z-score (P = .02) in the fully adjusted model (Table III). No associations were observed between total screen-time and lipids, BP, or clustered cardiovascular risk in any of the models.
Discussion Consistent with a number of previous cross-sectional studies in children,9,26-29 we found that TV viewing time was positively associated with adiposity. However, a unique aspect 4
of our study is the adjustment for objectively-measured MVPA and diet. That is, in the fully adjusted model there was a 0.25 increase in BMI z-score and 1.42 cm increase in WC for every hour of TV viewing. Although several previous studies have observed no association between TV viewing and adiposity,30-32 comparability with these studies is difficult because of a lack of adjustment for important covariates30,32 and differences in the measurement of TV viewing time (eg, time use diaries,31 self-reported time per day using a Likert scale,30 and self-reported number of days per week watching TV32). The positive association found between TV viewing and adiposity was not observed for e-games or using a computer. However, in the Transform-Us! cohort, Robinson et al
- 2015 time spent in these SBBs was small in comparison with time spent watching TV, and, therefore, the exposure may not have been sufficient to ascertain an association. Furthermore, computer use may not have the same relevance to adiposity, compared with TV viewing if it is not being used for recreation (eg, it is being used for homework). When all SBBs were combined as a measure of total daily screen-time in our study, we found no significant association with WC, and the relationship with BMI z-score was attenuated relative to TV viewing time. A similar result was reported in a study of 2- to 15-year-olds,33 whereby combining computer use with TV viewing was less strongly associated with BMI z-score compared with TV viewing alone. In our study, it is likely that the positive association between total screen-time and BMI z-score was driven by TV viewing time because daily time engaged in e-games and computer use accounted for less than one-third of total screen-time, and these behaviors were not associated with any measure of adiposity. It has been suggested that single markers, such as TV viewing, are unlikely to explain relationships between sedentary behavior and health.34 However, the results of our study suggest that some but not all SBBs are related to CVD risk factors. Therefore, the combination of SBBs as a measure of total screen-time may not represent an ideal measure when examining important health associations of specific behaviors. A unique finding from our study was the positive association between TV viewing time and SBP, independent of potential confounders. For every hour of TV viewing, there was a 1.52 mm Hg increase in SBP. Even though this finding is consistent with previous studies of TV viewing time and BP in children,8,10 comparability between studies is difficult because most failed to adjust for adiposity,10 and none adjusted for a measure of diet energy density or MVPA. In the EYHS of children and adolescents, no association was observed between TV viewing and SBP before or after adjusting for physical activity (average daily counts per minute) and adiposity.9 However, in that study TV viewing was only measured on weekdays, which may have resulted in an underestimation of total TV viewing time. In addition BP was only measured on one occasion, which limits the accuracy of the measurement used in analysis. It is also unclear why TV viewing, but no other sedentary SBBs, was associated with elevated SBP although higher levels of snacking while watching TV have been associated with elevated SBP and overall BP.8 Although we adjusted for diet energy density in our study, snacking during TV viewing was not specifically assessed and may be an important mediator that is relevant to TV viewing but not for other SBBs. In addition, similar to the positive association between TV viewing and adiposity that was not observed for other SBBs, it may be that the level of exposure was not sufficient to detect an association and that computer use may not have the same relevance to health if it is not being used for recreation. With the exception of a positive association between e-games and LDL-C, no associations between any of the
ORIGINAL ARTICLES SBBs and cholesterol or lipid concentrations were observed. This contrasts with the National Health and Nutrition Examination Survey (NHANES) study, which reported a positive association between TV viewing and non-HDL-C.11 However, because maturation influences lipid concentrations,20 the older age range included in that study and a longer lifetime exposure to sedentary behavior may account for the positive association. Furthermore, in comparison with NHANES, sedentary e-games were separated from active e-games in the Transform-Us! study and may account for the differences observed between studies. The finding that time spent playing e-games was positively associated with LDL-C could be explained by observations from previous research in which children who spent extended time playing e-games were significantly less active.35 Even though the adjustment for MVPA in our study suggests that this association was independent of activity levels, it may be that the time spent in light-intensity activities, which are beneficial to health, was lower among children who spent more time in sedentary e-games. When we assessed CVD risk with a composite score, we found no association with any of the SBBs. This is in contrast to the EYHS9 and NHANES,11 which reported positive associations between TV viewing and clustered metabolic risk. However, as previously described, MVPA and diet were not adjusted for, and BP is likely to have been overestimated in the EYHS. In spite of a broader age range, sexual maturation was not accounted for in NHANES. Furthermore, comparability in the outcome is limited because the components of the risk score differed between studies. The EYHS included insulin and glucose, whereas NHANES included C-reactive protein assessments. The inclusion of a marker of inflammation is suggested to have greater sensitivity and utility, compared with traditional risk factors.36 Therefore, the null association observed in the current study may be due to a lack of sensitivity of our measure. It may also reflect a relatively healthy sample with regard to the CVD risk factors measured and/or low levels of time spent playing e-games or using a computer. As computer use was not associated with any of the CVD risk factors and e-games was only associated with LDL-C, it is unlikely that there would be any evidence of clustering of CVD risk factors. A limitation is generalizability of the sample. Although schools were randomly selected, schools and children were able to opt in and children were required to provide active consent. Therefore, selection bias is a consideration with the possibility that healthier children were more likely to participate in the study. However, the sample had similar levels of compliance with screen-time guidelines in comparison with population levels in Australian children6 (41% in current study vs 38% of Australian children aged 5-8 years and 25% aged 9-11 years), but overweight/obesity levels were slightly lower (21% in current study vs 25% of Australian children). Generalizability may also be reduced by the lower percentage of children with active consent who provided a fasting morning blood sample (37%) compared with those
Screen-Based Behaviors of Children and Cardiovascular Risk Factors
5
THE JOURNAL OF PEDIATRICS
Vol. -, No. -
www.jpeds.com
who did not. However, although not all consenting children agreed to provide a fasting morning blood sample, it is important to note that there was no difference in the nonblood measures (SBP, DBP, and WC), time spent in each of the SBBs, or covariates (MVPA and parent education) between children who did and did not complete the blood sample (P > .05). Other limitations of this study include the use of a parental-proxy report to assess children’s screen-time and the cross-sectional design; thus, causality cannot be established. Public health strategies and interventions, which aim to reduce CVD risk factors in children, may, therefore, need to consider targeting different SBBs. However, further research is still needed to examine the effects of different types of SBBs on cardiometabolic health outcomes early in life. Given the changing nature of the technological environment, the assessment of additional types of screen media and the relation to CVD health may also be important. n
13.
14.
15.
16. 17.
18.
19. Submitted for publication Dec 20, 2014; last revision received Aug 4, 2015; accepted Aug 27, 2015. Reprint requests: Jo Salmon, PhD, Centre for Physical Activity and Nutrition Research, Deakin University, 221 Burwood Highway, Burwood, Victoria 3125, Australia. E-mail: [email protected]
20.
References
22.
1. Tremblay MS, LeBlanc AG, Janssen I, Kho ME, Hicks A, Murumets K, et al. Canadian sedentary behaviour guidelines for children and youth. Appl Physiol Nutr Metab 2011;36:59-64. 65-71. 2. American Academy of Pediatrics Committee on Public Education. Children, adolescents, and television. Pediatrics 2001;107:423-5. 3. American Academy of Pediatrics. Children, adolescents, and the media. Pediatrics 2013;132:958-61. 4. The Department of Health. Australia’s Physical Activity and Sedentary Behaviour Guidelines for Children (5–12 years), http://www.health. gov.au/internet/main/publishing.nsf/content/health-pubhlth-strategphys-act-guidelines#apa512; 2014. Accessed May 1, 2014. 5. Active Healthy Kids Canada. Are We Driving Our Kids to Unhealthy Habits? The 2013 Active Healthy Kids Canada Report Card on Physical Activity for Children and Youth, 2013. http://www.activehealthykids.ca/ ReportCard/2013ReportCard.aspx. Accessed September 11, 2013. 6. Australian Bureau of Statistics. Australian Health Survey: Updated Results, 2011-2012. http://www.abs.gov.au. Accessed September 11, 2013. 7. Sisson SB, Church TS, Martin CK, Tudor-Locke C, Smith SR, Bouchard C, et al. Profiles of sedentary behavior in children and adolescents: the US National Health and Nutrition Examination Survey, 20012006. Int J Pediatr Obes 2009;4:353-9. 8. Lazarou C, Panagiotakos DB, Matalas AL. Lifestyle factors are determinants of children’s blood pressure levels: the CYKIDS study. J Hum Hypertens 2009;23:456-63. 9. Ekelund U, Brage S, Froberg K, Harro M, Anderssen SA, Sardinha LB, et al. TV viewing and physical activity are independently associated with metabolic risk in children: the European Youth Heart Study. PLoS Med 2006;3:e488. 10. Martinez-Gomez D, Tucker J, Heelan KA, Welk GJ, Eisenmann JC. Associations between sedentary behavior and blood pressure in young children. Arch Pediatr Adolesc Med 2009;163:724-30. 11. Carson V, Janssen I. Volume, patterns, and types of sedentary behavior and cardio-metabolic health in children and adolescents: a crosssectional study. BMC Public Health 2011;11:274-83. 12. Salmon J, Arundell L, Hume C, Brown H, Hesketh K, Dunstan D, et al. A cluster-randomized controlled trial to reduce sedentary behavior and
23.
6
21.
24.
25.
26.
27.
28.
29.
30.
31.
32.
promote physical activity and health of 8-9 year olds: the TransformUs! Study. BMC Public Health 2011;11:759-67. Salmon J, Ball K, Hume C, Booth M, Crawford D. Outcomes of a grouprandomized trial to prevent excess weight gain, reduce screen behaviours and promote physical activity in 10-year-old children: Switch-Play. Int J Obes (Lond) 2008;32:601-12. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 2000;320:1240-3. Garnett SP, Cowell CT, Baur LA, Shrewsbury VA, Chan A, Crawford D, et al. Increasing central adiposity: the Nepean longitudinal study of young people aged 7-8 to 12-13 years. Int J Obes (Lond) 2005;29: 1353-60. Eisenmann JC. Waist circumference percentiles for 7- to 15-year-old Australian children. Acta Paediatr 2005;94:1182-5. Zimmet P, Alberti KG, Kaufman F, Tajima N, Silink M, Arslanian S, et al. The metabolic syndrome in children and adolescents - an IDF consensus report. Pediatr Diabetes 2007;8:299-306. National High Blood Pressure Education Program Working Group on High Blood Pressure in Children and Adolescents. The fourth report on the diagnosis, evaluation and treatment of high blood pressure in children and adolescents. Pediatrics 2004;114(Suppl 2):555-76. Department of Health and Human Services. 2000 CDC growth charts for the United States: methods and development. www.cdc.gov/ growthcharts. Accessed March 25, 2013. Daniels SR, Greer FR. Lipid screening and cardiovascular health in childhood. Pediatrics 2008;122:198-208. Campbell KJ, Crawford DA, Ball K. Family food environment and dietary behaviors likely to promote fatness in 5-6 year-old children. Int J Obes (Lond) 2006;30:1272-80. Freedson P, Pober D, Janz KF. Calibration of accelerometer output for children. Med Sci Sports Exerc 2005;37:S523-30. Trost SG, Loprinzi PD, Moore R, Pfeifer KA. Comparison of accelerometer cut points for predicting activity intensity in youth. Med Sci Sports Exerc 2011;43:1360-8. Cain KL, Sallis JF, Conway TL, Van Dyck D, Calhoon L. Using accelerometers in youth physical activity studies: a review of methods. J Phys Act Health 2013;10:437-50. Mattocks C, Ness A, Leary SD, Tilling K, Blair SN, Shield J, et al. Use of accelerometers in a large field based study of children: protocols, design issues, and effects on precision. J Phys Act Health 2008;5(Suppl 1):S98111. Crespo CJ, Smit E, Troiano RP, Bartlett SJ, Macera CA, Andersen RE. Television watching, energy intake, and obesity in US children: results from the third National Health and Nutrition Examination Survey, 1988-1994. Arch Pediatr Adolesc Med 2001;155:360-5. Dowda M, Ainsworth BE, Addy L, Saunders R, Riner W. Environmental influences, physical activity, and weight status in 8- to 16-year-olds. Arch Pediatr Adolesc Med 2001;155:711-7. Andersen RE, Crespo CJ, Bartlett SJ, Cheskin LJ, Pratt M. Relationship of physical activity and television watching with body weight and level of fatness among children - Results from the Third National Health and Nutrition Examination Survey. JAMA 1998;279:938-42. Hesketh K, Wake M, Graham M, Waters E. Stability of television viewing and electronic game/computer use in a prospective cohort study of Australian children: relationship with body mass index. Int J Behav Nutr Phys Act 2007;4:60-4. Fairclough SJ, Boddy LM, Hackett AF, Stratton G. Associations between children’s socioeconomic status, weight status and sex, with screenbased sedentary behaviours and sport participation. Int J Pediatr Obes 2009;4:299-305. Vandewater EA, Shim M, Caplovitz AG. Linking obesity and activity level with children’s television and video game use. J Adolesc 2004;27: 71-85. Graf C, Koch B, Kretschmann-Kandel E, Falkowski G, Christ H, Coburger S, et al. Correlation between BMI, leisure habits and motor abilities in childhood (CHILT-Project). Int J Obes Relat Metab Disord 2004;28:22-6.
Robinson et al
- 2015 33. Fulton JE, Wang X, Yore MM, Carlson SA, Galuska DA, Caspersen CJ. Television viewing, computer use, and BMI among U.S. children and adolescents. J Phys Act Health 2009;6(Suppl 1): S28-35. 34. Marshall SJ, Biddle SJH, Gorely T, Cameron N, Murdey I. Relationships between media use, body fatness and physical activity in children and youth: a meta-analysis. Int J Obes Relat Metab Disord 2004;28: 1238-46.
ORIGINAL ARTICLES 35. Salmon J, Timperio A, Telford A, Carver A, Crawford D. Association of family environment with children’s television viewing and with low level of physical activity. Obes Res 2005;13:1939-51. 36. Balagopal P, de Feranti SD, Cook S, Daniels SR, Gidding SS, Hayman LL, et al. Nontraditional risk factors and biomarkers for cardiovascular disease: mechanistic, research, and clinical considerations for youth: a scientific statement from the American Heart Association. Circulation 2011;123:2749-69.
Screen-Based Behaviors of Children and Cardiovascular Risk Factors
7
S
creen-based behaviors (SBBs; television [TV] viewing, computer use, and playing electronic games [e-games]) are key children’s sedentary leisure time pursuits. The American Academy of Pediatrics and government health authorities in the US, Australia, and Canada recommends limiting recreational screen-time to no more than 2 hours per day.1-4 However, population estimates in these countries show that between 40% and 80% of young people are exceeding screen-time recommendations.5-7 Since the initial establishment of pediatric screen-time recommendations in the US,2 which were based on time spent watching TV, pursuits such as using a computer and playing sedentary e-games have become highly pervasive among young people. Sedentary e-games and computer use may have physiological and metabolic effects that are not comparable with the more passive nature of watching TV and may have different behavioral mediators, such as snacking, which has been associated with cardiovascular disease (CVD) risk factors.8 To date, TV viewing and obesity are the dominant measures used in studies of screentime and CVD risk factors. A limited number of studies have examined the association between e-games and computer use with blood pressure (BP)8-10 and lipids9,11 in primary school children, and the findings are inconclusive. Few studies have accounted for the potential confounding effects of moderate-to-vigorous physical activity (MVPA) and diet (energy) quality, despite being independent predictors of CVD risk factors.9 Consequently, comparability between studies is restricted and the true CVD risk associated with different SBBs remains uncertain. Therefore, the aim of this study was to examine the independent associations between distinct SBBs (TV viewing, e-games, and computer use) and CVD risk factors and a clustered CVD risk score in a cohort of 7- to 10-year-old Australian children.
BMI BP CVD DBP e-games EYHS HDL-C LDL-C
Body mass index Blood pressure Cardiovascular disease Diastolic blood pressure Electronic games European Youth Heart Study High density lipoprotein cholesterol Low density lipoprotein cholesterol
MVPA NHANES SBB SBP SES TC TG TV WC
Moderate-to-vigorous physical activity National Health and Nutrition Examination Survey Screen-based behavior Systolic blood pressure Socioeconomic status Total cholesterol Triglycerides Television Waist circumference
From the Center for Physical Activity and Nutrition Research, School of Exercise and Nutrition Sciences, Deakin University, Burwood, Victoria, Australia Transform-Us! was supported by the National Health and Medical Research Council of Australia (2009-2013; ID533815) and the Diabetes Australia Research Trust. S.R. is supported by the National Heart Foundation (postgraduate scholarship). N.R. is supported by the Australian Research Council (Discovery Early Career Researcher Award DE120101173). J.S. is supported by the National Health and Medical Research Council (Principal Research Fellowship APP1026216). The authors declare no conflicts of interest. 0022-3476/$ - see front matter. Copyright ª 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpeds.2015.08.067
1
THE JOURNAL OF PEDIATRICS
www.jpeds.com
Methods Data were drawn from baseline of the Transform-Us! (ACTRN12609000715279; ISRCTN83725066) intervention trial, which has been described in detail elsewhere.12 In brief, Transform-Us! was a cluster-randomized controlled trial that aimed to reduce sedentary behavior, increase physical activity, and improve children’s health. Twenty primary (elementary) schools within a 50 km (31 miles) radius of the city of Melbourne, Australia, with an enrolment of over 300 students stratified by socioeconomic status (SES; 8 of 41 schools in low SES areas and 12 of 96 schools in mid-high SES areas) were randomly selected. All children in year 3 (n = 1606) at the participating schools were invited to take part. Fivehundred ninety-one children (37%) and 446 parents (28%) completed baseline assessments, which were used for the present study. Information was not obtained concerning nonresponders; it is an ethics requirement in Australia for active informed consent to be provided. A subsample of 344 children (21%) consented to provide a blood sample, and 219 of those children completed the assessment. Two-hundred sixty-four children had at least one CVD risk factor measure. Ethics approval was obtained from the Deakin University Human Research Ethics Committee, the Victorian Department of Education and Early Childhood Development, and the Catholic Education Office Melbourne. All parents were informed of the study purpose prior to randomization, therefore, any changes in children’s SBBs as a consequence of this information would have been consistent across the whole sample enabling the data to be pooled at baseline for analysis. The time spent in 3 types of sedentary SBBs: (1) TV viewing, videos, and DVDs; (2) nonactive e-games (eg, PlayStation, Nintendo) and computer games; and (3) computer and internet use (excluding games) were reported by the main parent or caregiver (referred to here-on in as the parent) of the child. Data concerning whether or not the child usually participated in a range of leisure activities during a typical week in the current school term (yes/no), the total hours and minutes from Monday to Friday, and the total hours and minutes on Saturday and Sunday were collected.13 Total screen-time (h/d) was calculated as the sum of time spent in TV viewing, e-games, and computer use. Height, weight, and waist circumference (WC) were measured by trained research staff at the child’s school. Height was measured without shoes to the nearest 0.1 cm using a portable stadiometer (Seca 220; Seca, Hamburg, Germany). Weight was measured in light clothing to the nearest 0.1 kg using portable electronic Tanita scales (Wedderburn 1582; Wedderburn, Sydney, Australia). Each was measured twice, and the average of the two measures determined. Body mass index (BMI) was calculated from the average of the height and weight measures as kg/m2 to classify overweight ($25 kg/ m2) and obesity ($30 kg/m2) according to International Obesity Task Force age- and sex-specific cut-points that correspond to BMI values in adults.14 WC was measured using a flexible steel tape at the narrowest point between the 2
Vol. -, No. bottom rib and the iliac crest, in the midaxillary plane over light clothing.15 Age- and sex-specific percentiles were derived using Australian specific data.16 A WC $90th percentile was used to classify overweight and obesity.17 Resting BP was measured in accordance with standard procedures and recommendations.18 After 2 minutes of seated quiet rest, BP was measured on the right arm with an automatic digital BP machine (Omron Hem-907; Omron, Melbourne, Australia) using an appropriately sized pediatric cuff. Three measurements were taken at 1-minute intervals on two occasions, one week apart (total of 6 readings). The first measurement was discarded from each visit, and the average was calculated from the remaining 4 measurements. Systolic (SBP) and diastolic BP (DBP) percentiles were determined using height percentiles for sex and age.19 Elevated BP was defined as SBP and/or DBP $90th and <95th percentile; hypertension was defined as SBP and/or DBP $95th percentile.18 Children were provided with Eutectic Mixture of Local Anaesthetics anesthetic cream and instructed to attend a local pathology clinic to provide a fasting morning blood sample. Total cholesterol (TC), high density lipoprotein cholesterol (HDL-C), low density lipoprotein cholesterol (LDL-C), and triglycerides (TG) were assessed using standard techniques at a National Association of Testing Authorities/Royal College of Pathologists Australasia accredited pathology laboratory. Ageand sex-specific percentile cut-points were used to determine children at risk: elevated TC, LDL-C, and TG were defined as $90th percentile and low HDL-C as #10th percentile.20 Based on previous cardiometabolic risk score methodology from the European Youth Heart Study (EYHS),9 a continuous clustered CVD risk score was derived using the following 6 variables: SBP, DBP, LDL-C, HDL-C, TG, and WC. The calculation for average SBP and average DBP are described above, and these variables were used in the calculation of the CVD risk score. As HDL-C is inversely related to cardiovascular risk it was multiplied by 1 to indicate higher CVD risk with increasing value. For each of the 6 variables, the z-score was calculated as the number of SD units from the sample mean after normalization of the variable (Z = [value mean]/SD). All standardized scores were summed to create a clustered CVD risk score with a higher score indicating a higher level of risk. An estimate of children’s consumption of 6 key energy dense food types (salty snacks, chocolate and sweets, cakes, pastries, fast food, chips) and two beverage types (fruit juice and soft drinks) was obtained from a parental-proxy report. These 8 items have previously been identified as important contributors to children’s energy intake.21 Responses for food items were assessed on a monthly 9-point scale ranging from “never or less than once a month” (score = 1) to “6 or more times a day” (score = 9). Responses for beverage items were assessed on a daily 8-point scale ranging from “my child does not drink this beverage” (score = 1) to “6 or more servings a day” (score = 8). The frequency for each item was summed to provide a total diet energy density score with a higher score indicating greater energy density of the diet.21 The responding parent was also asked to select their highest Robinson et al
- 2015
ORIGINAL ARTICLES
level of education (never attended school to university or tertiary qualification). Children wore an ActiGraph GT3X accelerometer (ActiGraph, Pensacola, Florida) for 8 consecutive days on a belt positioned on the right hip during waking hours, except during water-based activities. The normal frequency filter was selected, and the epoch length was 15 seconds. MVPA (min/d) was defined using age-adjusted cut-points22 defined as $4 METs.23 Nonwear time is commonly defined in the literature as 20 minutes of consecutive zeros, and this definition was used in the current study.24 Wear time criterion was defined as 3 valid weekdays ($8 hours) and 1 valid weekend day ($7 hours).25 Data Analyses All analyses were undertaken using Stata/SE v 12.0 (StataCorp, College Station, Texas). The distributions of each of the outcome variables were assessed for normality. Because BMI was positively skewed, age- and sex-adjusted BMI zscores14 were calculated and used in all analyses. Descriptive data (mean, SD) were calculated for CVD risk factors. There were no differences in the anthropometric and SBP or DBP measures between children who provided a blood sample and those who did not (data not shown); therefore, all children with these health data were included in the analyses. Forced entry linear regression analyses were used to examine the association between daily TV viewing, e-games, computer use, and total screen-time (h/d) and CVD risk factors. Four statistical models were used to explore the associations and adjust for important confounding factors: model 1 adjusted for sex, school clustering, and parent education; model 2 additionally adjusted for MVPA and accelerometer wear time; model 3 additionally adjusted for diet energy density; and model 4 additionally adjusted for WC (where adiposity was not the outcome of interest). Only children with data for all covariates were included in analyses (Table I).
Results Characteristics and the number of children (enrolled February to June 2010) providing complete data for each variable are shown in Table I. The mean age was 8.7 (0.4) years, and approximately one-half of the parents had completed a university degree. Nearly 60% of children exceeded screentime recommendations of >2 hours per day. Approximately 1 in 5 children were classified as overweight or obese based on BMI (Table II). A high percentage of children had elevated TC levels (37.5%); however, only 2 children had HDL-C below the 10th percentile. Seventeen percent of the children had a SBP or DBP greater than the 90th percentile. TV Viewing TV viewing time was positively associated with BMI z-score (P = .002) and WC (P = .02) after adjusting for sex of the child, parent education, MVPA, accelerometer wear time, and diet
Table I. Participant characteristics Measures
Valid (n)
Mean (SD)
Age (y) Parent education (% university) Accelerometer MVPA (min/d) Accelerometer wear time (min/d) SBBs TV (min/d) E-games (min/d) Computer (min/d) Total screen (min/d) Screen-time >2 h/d (%) Diet Diet composite score Anthropometry Height (cm) Weight (kg) BMI (kg/m2) BMI z-score WC (cm) BP SBP (mm Hg) DBP (mm Hg) Cholesterol and lipids TC (mmol/L) HDL-C (mmol/L) LDL-C (mmol/L) TG (mmol/L) Clustered risk score*
264 264
8.7 (0.4) 46.2
264 264
66.9 (19.5) 699.0 (53.5)
262 264 263 261 261
101.5 (56.1) 25.4 (32.9) 19.6 (21.8) 146.4 (78.9) 58.6
264
21.0 (5.1)
263 263 263 263 264
132.0 (6.5) 30.0 (6.1) 17.1 (2.5) 0.2 (0.9) 59.1 (6.3)
255 255
101.7 (9.2) 60.1 (8.02)
147 147 147 147 147
4.59 (0.78) 1.69 (0.33) 2.58 (0.73) 0.69 (0.25) 0.29 (2.7)
All values represent mean (SD) or proportion (percentage). *Standardized and sum of z-scores for WC, the average of SBP and DBP, LDL-C, inverted HDLC, and TG.
energy density (Table III). A significant positive association was also found between TV viewing time and SBP (P = .05) after adjusting for sex, parent education, MVPA, accelerometer wear time, diet energy density, and WC. E-Games Time spent playing e-games was positively associated with LDL-C (P = .05) in the fully adjusted model. There was no association between e-games and any of the other cholesterol and lipid measures, BP, adiposity, or clustered CVD risk in any of the models. Computer Use Time spent using a computer was not associated with any of the CVD risk factors or clustered CVD risk scores in any of the models. Table II. Percentage of children with elevated levels of CVD risk factors CVD risk factors 2
BMI (>25 kg/m ) WC $75th percentile WC $90th percentile DBP and/or SBP >90th to <95th percentile DBP and/or SBP $95th percentile TC >90th percentile HDL-C <10th percentile LDL-C >90th percentile TG $90th percentile
Screen-Based Behaviors of Children and Cardiovascular Risk Factors
Valid (n)
%
263 264 264 255 255 147 147 147 147
22.1 30.7 12.9 6.6 10.5 37.5 1.4 12.9 19.7 3
THE JOURNAL OF PEDIATRICS
Vol. -, No. -
www.jpeds.com
Table III. Unstandardized regression coefficients of time (h/d) spent in SBBs and CVD risk factors TV
Adiposity (n = 261) z-BMI (kg/m2) Model 1 Model 2 Model 3 WC (cm) Model 1 Model 2 Model 3 BP (n = 255) SBP (mm Hg) Model 1 Model 2 Model 3 Model 4 DBP (mm Hg) Model 1 Model 2 Model 3 Model 4 Cholesterol and lipids (n = 146) HDL-C (mmol/L) Model 1 Model 2 Model 3 Model 4 LDL-C (mmol/L) Model 1 Model 2 Model 3 Model 4 TC (mmol/L) Model 1 Model 2 Model 3 Model 4 TG (mmol/L) Model 1 Model 2 Model 3 Model 4 CVD risk score* Model 1 Model 2 Model 3
E-games b (95% CI)
Computer P
b (95% CI)
Total screen
b (95% CI)
P
P
b (95% CI)
P
0.22 (0.08, 0.34) 0.22 (0.09, 0.36) 0.25 (0.10, 0.39)
.004 .003 .002
0.00 ( 0.29, 0.28) 0.01 ( 0.28, 0.26) 0.00 ( 0.29, 0.28)
.97 .92 .99
0.18 ( 0.50, 0.14) 0.18 ( 0.49, 0.14) 0.17 ( 0.49, 0.14)
.26 .26 .26
0.10 (0.01, 0.19) 0.10 (0.01, 0.19) 0.12 (0.02, 0.21)
.04 .04 .02
1.18 (0.09, 2.26) 1.24 (0.14, 2.34) 1.42 (0.29, 2.55)
.04 .03 .02
0.05 ( 2.08, 1.98) 0.16 ( 2.06, 1.73) 0.01 ( 2.12, 2.15)
.96 .86 .99
0.74 ( 3.08, 1.61) 0.67 ( 2.96, 1.62) 0.61 ( 2.84, 1.61)
.52 .55 .57
0.57 ( 0.20, 1.33) 0.58 ( 0.16, 1.33) 0.71 ( 0.08, 1.51)
.14 .12 .07
1.82 (0.23, 3.40) 1.84 (0.23, 3.45) 1.83 (0.18, 3.49) 1.52 (0.00, 3.04)
.03 .03 .03 .05
1.22 ( 1.32 ( 1.71 ( 1.74 (
3.72, 1.27) 3.86, 1.20) 4.14, 0.72) 4.04, 0.56)
.32 .29 .16 .13
1.76 ( 1.79 ( 1.69 ( 1.88 (
2.10, 5.63) 2.17, 5.74) 2.16, 5.55) 1.90, 5.67)
.35 .36 .37 .31
0.86 ( 0.85 ( 0.81 ( 0.60 (
0.36, 2.07) 0.37, 2.07) 0.41, 2.04) 0.58, 1.79)
.16 .16 .18 .30
0.54 ( 0.63 ( 0.60 ( 0.38 (
0.78, 1.87) 0.77, 2.03) 0.92, 2.11) 1.14, 1.91)
.40 .36 .42 .60
0.55 ( 0.89 ( 1.08 ( 1.09 (
3.18, 2.07) 3.32, 1.54) 3.59, 1.44) 3.66, 1.47)
.66 .45 .38 .39
0.80 ( 1.95, 3.54) 0.91 ( 1.90, 3.72) 0.87 (1.96, 3.71) 0.99 ( 1.81, 3.78)
.55 .51 .53 .47
0.28 ( 0.27 ( 0.23 ( 0.10 (
0.76, 1.31) 0.79, 1.33) 0.91, 1.38) 1.07, 1.28)
.59 .60 .68 .86
0.00 ( 0.01 ( 0.01 ( 0.03 (
0.06, 0.06) 0.06, 0.06) 0.06, 0.05) 0.02, 0.07)
.96 .86 .83 .31
0.06 ( 0.04 ( 0.04 ( 0.03 (
0.14, 0.01) 0.11, 0.02) 0.12, 0.03) 0.10, 0.04)
.10 .15 .19 .35
0.10 ( 0.09 ( 0.09 ( 0.06 (
0.08, 0.28) 0.12, 0.29) 0.11, 0.29) 0.13, 0.24)
.25 .38 .38 .52
0.00 ( 0.00 ( 0.00 ( 0.01 (
0.39, 0.04) 0.04, 0.04) 0.04, 0.04) 0.02, 0.05)
.97 .96 .93 .42
0.01 ( 0.02 ( 0.03 ( 0.00 (
0.12, 0.13) 0.10, 0.13) 0.08, 0.14) 0.10, 0.10)
.92 .76 .57 .99
0.12 ( 0.02, 0.27) 0.09 ( 0.04, 0.24) 0.13 (0.01, 0.25) 0.11 (0.00, 0.22)
.09 .16 .04 .05
0.04 ( 0.01 ( 0.00 ( 0.04 (
0.35, 0.28) 0.32, 0.30) 0.32, 0.33) 0.28, 0.35)
.81 .94 .98 .81
0.02 ( 0.02 ( 0.04 ( 0.02 (
0.06, 0.10) 0.05, 0.10) 0.04, 0.11) 0.05, 0.09)
.57 .49 .29 .49
0.02 ( 0.02 ( 0.04 ( 0.05 (
0.09, 0.12) 0.08, 0.13) 0.05, 0.13) 0.04, 0.13)
.72 .62 .39 .26
0.05 ( 0.03 ( 0.05 ( 0.04 (
0.10, 0.20) 0.12, 0.17) 0.09, 0.18) 0.11, 0.18)
.52 .72 .47 .58
0.04 ( 0.07 ( 0.09 ( 0.09 (
0.31, 0.40) 0.27, 0.41) 0.26, 0.43) 0.27, 0.44)
.80 .68 .61 .62
0.03 ( 0.03 ( 0.04 ( 0.05 (
0.05, 0.10) 0.04, 0.10) 0.02, 0.11) 0.01, 0.11)
.44 .35 .19 .14
0.01 ( 0.00 ( 0.00 ( 0.01 (
0.06, 0.04) 0.05, 0.04) 0.04, 0.04) 0.06, 0.03)
.61 .81 .99 .58
0.05 ( 0.04 ( 0.05 ( 0.04 (
0.03, 0.14) 0.06, 0.13) 0.05, 0.14) 0.05, 0.13)
.17 .43 .31 .37
0.07 ( 0.06 ( 0.05 ( 0.04 (
0.18, 0.04) 0.18, 0.07) 0.17, 0.07) 0.16, 0.07)
.19 .34 .38 .46
0.00 ( 0.00 ( 0.00 ( 0.00 (
0.04, 0.03) 0.03, 0.03) 0.03, 0.03) 0.31, 0.03)
.84 .91 .83 .84
0.28 ( 0.35, 0.91) 0.37 ( 0.23, 0.96) 0.46 ( 0.17, 1.10)
.36 .21 .14
0.61 ( 0.24, 1.45) 0.38 ( 0.36, 1.12) 0.55 ( 0.18, 1.29)
.15 .30 .13
0.86 ( 2.24, 0.51) 0.66 ( 2.14, 0.81) 0.58 ( 2.06, 0.89)
.20 .36 .42
0.17 ( 0.25, 0.59) 0.19 ( 0.21, 0.59) 0.28 ( 0.16, 0.71)
.41 .33 .19
b, beta coefficient. Model 1 adjusted for school clustering, SES, and sex of the child: model 2, additionally adjusted for MVPA and accelerometer wear time; model 3, additionally adjusted for a diet composite score; model 4, additionally adjusted for WC. Bold type indicates significance (P < .05). *Standardized and sum of WC, the average of SBP and DBP, LDL-C, inverted HDL-C, and TG.
Total Screen-Time Average daily total screen-time was positively associated with BMI z-score (P = .02) in the fully adjusted model (Table III). No associations were observed between total screen-time and lipids, BP, or clustered cardiovascular risk in any of the models.
Discussion Consistent with a number of previous cross-sectional studies in children,9,26-29 we found that TV viewing time was positively associated with adiposity. However, a unique aspect 4
of our study is the adjustment for objectively-measured MVPA and diet. That is, in the fully adjusted model there was a 0.25 increase in BMI z-score and 1.42 cm increase in WC for every hour of TV viewing. Although several previous studies have observed no association between TV viewing and adiposity,30-32 comparability with these studies is difficult because of a lack of adjustment for important covariates30,32 and differences in the measurement of TV viewing time (eg, time use diaries,31 self-reported time per day using a Likert scale,30 and self-reported number of days per week watching TV32). The positive association found between TV viewing and adiposity was not observed for e-games or using a computer. However, in the Transform-Us! cohort, Robinson et al
- 2015 time spent in these SBBs was small in comparison with time spent watching TV, and, therefore, the exposure may not have been sufficient to ascertain an association. Furthermore, computer use may not have the same relevance to adiposity, compared with TV viewing if it is not being used for recreation (eg, it is being used for homework). When all SBBs were combined as a measure of total daily screen-time in our study, we found no significant association with WC, and the relationship with BMI z-score was attenuated relative to TV viewing time. A similar result was reported in a study of 2- to 15-year-olds,33 whereby combining computer use with TV viewing was less strongly associated with BMI z-score compared with TV viewing alone. In our study, it is likely that the positive association between total screen-time and BMI z-score was driven by TV viewing time because daily time engaged in e-games and computer use accounted for less than one-third of total screen-time, and these behaviors were not associated with any measure of adiposity. It has been suggested that single markers, such as TV viewing, are unlikely to explain relationships between sedentary behavior and health.34 However, the results of our study suggest that some but not all SBBs are related to CVD risk factors. Therefore, the combination of SBBs as a measure of total screen-time may not represent an ideal measure when examining important health associations of specific behaviors. A unique finding from our study was the positive association between TV viewing time and SBP, independent of potential confounders. For every hour of TV viewing, there was a 1.52 mm Hg increase in SBP. Even though this finding is consistent with previous studies of TV viewing time and BP in children,8,10 comparability between studies is difficult because most failed to adjust for adiposity,10 and none adjusted for a measure of diet energy density or MVPA. In the EYHS of children and adolescents, no association was observed between TV viewing and SBP before or after adjusting for physical activity (average daily counts per minute) and adiposity.9 However, in that study TV viewing was only measured on weekdays, which may have resulted in an underestimation of total TV viewing time. In addition BP was only measured on one occasion, which limits the accuracy of the measurement used in analysis. It is also unclear why TV viewing, but no other sedentary SBBs, was associated with elevated SBP although higher levels of snacking while watching TV have been associated with elevated SBP and overall BP.8 Although we adjusted for diet energy density in our study, snacking during TV viewing was not specifically assessed and may be an important mediator that is relevant to TV viewing but not for other SBBs. In addition, similar to the positive association between TV viewing and adiposity that was not observed for other SBBs, it may be that the level of exposure was not sufficient to detect an association and that computer use may not have the same relevance to health if it is not being used for recreation. With the exception of a positive association between e-games and LDL-C, no associations between any of the
ORIGINAL ARTICLES SBBs and cholesterol or lipid concentrations were observed. This contrasts with the National Health and Nutrition Examination Survey (NHANES) study, which reported a positive association between TV viewing and non-HDL-C.11 However, because maturation influences lipid concentrations,20 the older age range included in that study and a longer lifetime exposure to sedentary behavior may account for the positive association. Furthermore, in comparison with NHANES, sedentary e-games were separated from active e-games in the Transform-Us! study and may account for the differences observed between studies. The finding that time spent playing e-games was positively associated with LDL-C could be explained by observations from previous research in which children who spent extended time playing e-games were significantly less active.35 Even though the adjustment for MVPA in our study suggests that this association was independent of activity levels, it may be that the time spent in light-intensity activities, which are beneficial to health, was lower among children who spent more time in sedentary e-games. When we assessed CVD risk with a composite score, we found no association with any of the SBBs. This is in contrast to the EYHS9 and NHANES,11 which reported positive associations between TV viewing and clustered metabolic risk. However, as previously described, MVPA and diet were not adjusted for, and BP is likely to have been overestimated in the EYHS. In spite of a broader age range, sexual maturation was not accounted for in NHANES. Furthermore, comparability in the outcome is limited because the components of the risk score differed between studies. The EYHS included insulin and glucose, whereas NHANES included C-reactive protein assessments. The inclusion of a marker of inflammation is suggested to have greater sensitivity and utility, compared with traditional risk factors.36 Therefore, the null association observed in the current study may be due to a lack of sensitivity of our measure. It may also reflect a relatively healthy sample with regard to the CVD risk factors measured and/or low levels of time spent playing e-games or using a computer. As computer use was not associated with any of the CVD risk factors and e-games was only associated with LDL-C, it is unlikely that there would be any evidence of clustering of CVD risk factors. A limitation is generalizability of the sample. Although schools were randomly selected, schools and children were able to opt in and children were required to provide active consent. Therefore, selection bias is a consideration with the possibility that healthier children were more likely to participate in the study. However, the sample had similar levels of compliance with screen-time guidelines in comparison with population levels in Australian children6 (41% in current study vs 38% of Australian children aged 5-8 years and 25% aged 9-11 years), but overweight/obesity levels were slightly lower (21% in current study vs 25% of Australian children). Generalizability may also be reduced by the lower percentage of children with active consent who provided a fasting morning blood sample (37%) compared with those
Screen-Based Behaviors of Children and Cardiovascular Risk Factors
5
THE JOURNAL OF PEDIATRICS
Vol. -, No. -
www.jpeds.com
who did not. However, although not all consenting children agreed to provide a fasting morning blood sample, it is important to note that there was no difference in the nonblood measures (SBP, DBP, and WC), time spent in each of the SBBs, or covariates (MVPA and parent education) between children who did and did not complete the blood sample (P > .05). Other limitations of this study include the use of a parental-proxy report to assess children’s screen-time and the cross-sectional design; thus, causality cannot be established. Public health strategies and interventions, which aim to reduce CVD risk factors in children, may, therefore, need to consider targeting different SBBs. However, further research is still needed to examine the effects of different types of SBBs on cardiometabolic health outcomes early in life. Given the changing nature of the technological environment, the assessment of additional types of screen media and the relation to CVD health may also be important. n
13.
14.
15.
16. 17.
18.
19. Submitted for publication Dec 20, 2014; last revision received Aug 4, 2015; accepted Aug 27, 2015. Reprint requests: Jo Salmon, PhD, Centre for Physical Activity and Nutrition Research, Deakin University, 221 Burwood Highway, Burwood, Victoria 3125, Australia. E-mail: [email protected]
20.
References
22.
1. Tremblay MS, LeBlanc AG, Janssen I, Kho ME, Hicks A, Murumets K, et al. Canadian sedentary behaviour guidelines for children and youth. Appl Physiol Nutr Metab 2011;36:59-64. 65-71. 2. American Academy of Pediatrics Committee on Public Education. Children, adolescents, and television. Pediatrics 2001;107:423-5. 3. American Academy of Pediatrics. Children, adolescents, and the media. Pediatrics 2013;132:958-61. 4. The Department of Health. Australia’s Physical Activity and Sedentary Behaviour Guidelines for Children (5–12 years), http://www.health. gov.au/internet/main/publishing.nsf/content/health-pubhlth-strategphys-act-guidelines#apa512; 2014. Accessed May 1, 2014. 5. Active Healthy Kids Canada. Are We Driving Our Kids to Unhealthy Habits? The 2013 Active Healthy Kids Canada Report Card on Physical Activity for Children and Youth, 2013. http://www.activehealthykids.ca/ ReportCard/2013ReportCard.aspx. Accessed September 11, 2013. 6. Australian Bureau of Statistics. Australian Health Survey: Updated Results, 2011-2012. http://www.abs.gov.au. Accessed September 11, 2013. 7. Sisson SB, Church TS, Martin CK, Tudor-Locke C, Smith SR, Bouchard C, et al. Profiles of sedentary behavior in children and adolescents: the US National Health and Nutrition Examination Survey, 20012006. Int J Pediatr Obes 2009;4:353-9. 8. Lazarou C, Panagiotakos DB, Matalas AL. Lifestyle factors are determinants of children’s blood pressure levels: the CYKIDS study. J Hum Hypertens 2009;23:456-63. 9. Ekelund U, Brage S, Froberg K, Harro M, Anderssen SA, Sardinha LB, et al. TV viewing and physical activity are independently associated with metabolic risk in children: the European Youth Heart Study. PLoS Med 2006;3:e488. 10. Martinez-Gomez D, Tucker J, Heelan KA, Welk GJ, Eisenmann JC. Associations between sedentary behavior and blood pressure in young children. Arch Pediatr Adolesc Med 2009;163:724-30. 11. Carson V, Janssen I. Volume, patterns, and types of sedentary behavior and cardio-metabolic health in children and adolescents: a crosssectional study. BMC Public Health 2011;11:274-83. 12. Salmon J, Arundell L, Hume C, Brown H, Hesketh K, Dunstan D, et al. A cluster-randomized controlled trial to reduce sedentary behavior and
23.
6
21.
24.
25.
26.
27.
28.
29.
30.
31.
32.
promote physical activity and health of 8-9 year olds: the TransformUs! Study. BMC Public Health 2011;11:759-67. Salmon J, Ball K, Hume C, Booth M, Crawford D. Outcomes of a grouprandomized trial to prevent excess weight gain, reduce screen behaviours and promote physical activity in 10-year-old children: Switch-Play. Int J Obes (Lond) 2008;32:601-12. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 2000;320:1240-3. Garnett SP, Cowell CT, Baur LA, Shrewsbury VA, Chan A, Crawford D, et al. Increasing central adiposity: the Nepean longitudinal study of young people aged 7-8 to 12-13 years. Int J Obes (Lond) 2005;29: 1353-60. Eisenmann JC. Waist circumference percentiles for 7- to 15-year-old Australian children. Acta Paediatr 2005;94:1182-5. Zimmet P, Alberti KG, Kaufman F, Tajima N, Silink M, Arslanian S, et al. The metabolic syndrome in children and adolescents - an IDF consensus report. Pediatr Diabetes 2007;8:299-306. National High Blood Pressure Education Program Working Group on High Blood Pressure in Children and Adolescents. The fourth report on the diagnosis, evaluation and treatment of high blood pressure in children and adolescents. Pediatrics 2004;114(Suppl 2):555-76. Department of Health and Human Services. 2000 CDC growth charts for the United States: methods and development. www.cdc.gov/ growthcharts. Accessed March 25, 2013. Daniels SR, Greer FR. Lipid screening and cardiovascular health in childhood. Pediatrics 2008;122:198-208. Campbell KJ, Crawford DA, Ball K. Family food environment and dietary behaviors likely to promote fatness in 5-6 year-old children. Int J Obes (Lond) 2006;30:1272-80. Freedson P, Pober D, Janz KF. Calibration of accelerometer output for children. Med Sci Sports Exerc 2005;37:S523-30. Trost SG, Loprinzi PD, Moore R, Pfeifer KA. Comparison of accelerometer cut points for predicting activity intensity in youth. Med Sci Sports Exerc 2011;43:1360-8. Cain KL, Sallis JF, Conway TL, Van Dyck D, Calhoon L. Using accelerometers in youth physical activity studies: a review of methods. J Phys Act Health 2013;10:437-50. Mattocks C, Ness A, Leary SD, Tilling K, Blair SN, Shield J, et al. Use of accelerometers in a large field based study of children: protocols, design issues, and effects on precision. J Phys Act Health 2008;5(Suppl 1):S98111. Crespo CJ, Smit E, Troiano RP, Bartlett SJ, Macera CA, Andersen RE. Television watching, energy intake, and obesity in US children: results from the third National Health and Nutrition Examination Survey, 1988-1994. Arch Pediatr Adolesc Med 2001;155:360-5. Dowda M, Ainsworth BE, Addy L, Saunders R, Riner W. Environmental influences, physical activity, and weight status in 8- to 16-year-olds. Arch Pediatr Adolesc Med 2001;155:711-7. Andersen RE, Crespo CJ, Bartlett SJ, Cheskin LJ, Pratt M. Relationship of physical activity and television watching with body weight and level of fatness among children - Results from the Third National Health and Nutrition Examination Survey. JAMA 1998;279:938-42. Hesketh K, Wake M, Graham M, Waters E. Stability of television viewing and electronic game/computer use in a prospective cohort study of Australian children: relationship with body mass index. Int J Behav Nutr Phys Act 2007;4:60-4. Fairclough SJ, Boddy LM, Hackett AF, Stratton G. Associations between children’s socioeconomic status, weight status and sex, with screenbased sedentary behaviours and sport participation. Int J Pediatr Obes 2009;4:299-305. Vandewater EA, Shim M, Caplovitz AG. Linking obesity and activity level with children’s television and video game use. J Adolesc 2004;27: 71-85. Graf C, Koch B, Kretschmann-Kandel E, Falkowski G, Christ H, Coburger S, et al. Correlation between BMI, leisure habits and motor abilities in childhood (CHILT-Project). Int J Obes Relat Metab Disord 2004;28:22-6.
Robinson et al
- 2015 33. Fulton JE, Wang X, Yore MM, Carlson SA, Galuska DA, Caspersen CJ. Television viewing, computer use, and BMI among U.S. children and adolescents. J Phys Act Health 2009;6(Suppl 1): S28-35. 34. Marshall SJ, Biddle SJH, Gorely T, Cameron N, Murdey I. Relationships between media use, body fatness and physical activity in children and youth: a meta-analysis. Int J Obes Relat Metab Disord 2004;28: 1238-46.
ORIGINAL ARTICLES 35. Salmon J, Timperio A, Telford A, Carver A, Crawford D. Association of family environment with children’s television viewing and with low level of physical activity. Obes Res 2005;13:1939-51. 36. Balagopal P, de Feranti SD, Cook S, Daniels SR, Gidding SS, Hayman LL, et al. Nontraditional risk factors and biomarkers for cardiovascular disease: mechanistic, research, and clinical considerations for youth: a scientific statement from the American Heart Association. Circulation 2011;123:2749-69.
Screen-Based Behaviors of Children and Cardiovascular Risk Factors
7