You are on page 1of 10

Atkin et al.

International Journal of Behavioral Nutrition and Physical Activity 2013, 10:137

RESEARCH Open Access

Bedroom media, sedentary time and screen-time
in children: a longitudinal analysis
Andrew J Atkin1*, Kirsten Corder1,2 and Esther M F van Sluijs1,2

Background: Having electronic media in the bedroom is cross-sectionally associated with greater screen-time in
children, but few longitudinal studies exist. The aim of this study was to describe longitudinal patterns of ownership
and examine cross-sectional and longitudinal associations of bedroom media with children’s sedentary behaviour.
Methods: Data are from the Sport, Physical activity and Eating behaviour: Environmental Determinants in Young
people (SPEEDY) study, collected at 3 time-points: baseline (2007, T0; age 10.3 ± 0.3 years), 1-year (T1y) and 4-year
(T4y) follow-up. For each assessment, 1512 (44.9% male), 715 (41.0% male), and 319 (48.3% male) participants
provided valid accelerometer data. Outcome variables were accelerometer-assessed sedentary time and
self-reported screen-time. The presence of a television or computer in the bedroom was self-reported by
participants and a combined bedroom media score calculated as the sum of such items. Cross-sectional and
longitudinal associations between bedroom media and each outcome were examined using multi-level linear
Results: Bedroom TV ownership fell from 70.9% at T0 to 42.5% at T4y. Having a TV in the bedroom
(beta; 95% CI*100, T0: -1.17; -1.88, -0.46. T1y: -1.68; -2.67, -0.70) and combined bedroom media (T0: -0.76; -1.26, -0.27.
T1y: -0.79; -1.51, -0.07) were negatively associated with objectively measured weekly sedentary time at T0 and T1y.
Having a computer in the bedroom (beta; 95% CI, T0: 0.15; 0.02, 0.29. T4y: 0.35; 0.10, 0.60) and combined bedroom
media (T0: 0.09: 0.01, 0.18. T4y: 0.20; 0.05, 0.34) were positively associated with screen-time at T0 and T4y. Relative to
participants without a computer throughout the study, children that had a computer in their bedroom at T0 but
not at T4y (beta; 95% CI for change in screen-time: -8.02; -12.75, -3.29) reported smaller increases in screen-time.
Conclusions: The bedroom media environment changes with age and exhibits a complex relationship with
children’s sedentary behaviour. Modifying children’s bedroom media environment may impact upon screen-time
but appears unlikely to influence overall sedentary time.
Keywords: Television, Children, Family, Sedentary behaviour, Accelerometer

Introduction that children should minimise the amount of time spent
Sedentary behaviours, such as watching television (TV) being sedentary for prolonged periods [11,12].
and using a computer, are highly prevalent during child- Research into the determinants of sedentary behaviour
hood [1-4] and may be adversely associated with cardio- enables the identification of at-risk populations and modifi-
metabolic health, though evidence from longitudinal and able factors that may be targeted within intervention
experimental research is limited [5-10]. In the UK, and programmes [13,14]. Contemporary thinking on the deter-
other countries, public health guidelines recommend minants of health behaviour advocates the application of an
ecological framework to reflect the influence of factors op-
erating at individual, social and environmental levels [15].
* Correspondence: In children, the influence of home and familial characteris-
UKCRC Centre for Diet and Activity Research (CEDAR), MRC Epidemiology tics on sedentary behaviour patterns has been a key area of
Unit, University of Cambridge School of Clinical Medicine, Institute of research [16-19], particularly regarding the impact of elec-
Metabolic Science, Cambridge Biomedical Campus, Box 285, Cambridge CB2
0QQ, UK tronic media (TV, computers, video games consoles) in the
Full list of author information is available at the end of the article

© 2013 Atkin et al.; licensee BioMed Central Ltd. This is an open access article distributed under the terms of the Creative
Commons Attribution License (, which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly cited.

all children to define sedentary time [34. using a computer). sedentary behaviour and dietary patterns in children Children were instructed to wear the monitors during from the county of Norfolk. Consent forms were room media with objectively measured sedentary time and returned to the study office by bedroom.mrc-epid. To increase recruitment. and 4- year follow-up (T4y. at least five original Previous research a valid home address from T1y (n = 1964) were sent in- examining the association between presence of a TV in the formation sheets and consent forms.35]. holiday (July 2011). Those who consented the bedroom has been recommended [9. UK [31]. from ages 9/10 to 13/14 years (3 waves of assess. using an ad- the bedroom media environment are associated with dressed. At participating schools.21] and this has later. However. Pensacola. reduced sleep and increased risk of study information sheets and consent forms were mailed overweight [20. for baseline data collection in SPEEDY have been re- http://www. At baseline.33] set to record at 5-second epochs. was required for inclusion in the analysis. Study design and setting The Sport. TV.25-27]. The removal of such devices from to all 2064 initial participants [2]. but our original consent did not allow us to to (1) describe changes in bedroom media ownership over trace individual participants. Participants questionnaire.2]. trained research assistants vis. pre-paid envelope. an ownership with changes in objectively measured sedentary extra invitation letter was sent home prior to the time and self-reported screen-time. [32. April-July 2007). Accelerometer data were ana- Full details of participant recruitment and procedures lysed using a batch processing program (MAHUffe. International Journal of Behavioral Nutrition and Physical Activity 2013. including 1 weekend day. We presented the study 4 years. Physical Activity. Through local bedroom and total sedentary time has been conducted in administrative authorities. To optimise specificity between outcome Data collection procedures and exposure measures. administer using a slightly modified version of a child self-report child questionnaires. 10:137 Page 2 of 10 http://www. age 10/11y.ijbnpa. participants also received an accelerometer diary and instructed to indicate when the monitor was taken Participants off and for what reason. FL) ac- investigating factors associated with physical activity. At T1y. Subsequent mea- self-reported screen-time at three time-points and (3) surements were taken at school following similar pro- examine the association of changes in bedroom media cedures as at baseline. with >12 children in year 5). poses.36] and days with <500 mi- an invitation pack.22-24]. leisure-time screen-time was assessed ited schools to take physical measurements. sedentary time was measured ob- study is a population based longitudinal. timing of follow-up assessments were matched Self-reported screen-time to baseline. cohort study jectively using an Actigraph (GT1M. Ethical approval waking hours for 7 days and to remove them while bath- was obtained from the University of East Anglia research ing. celerometer. April-July 2011). data collected during school Participants were invited to participate on three separate hours (9 am-3 pm on weekdays) were excluded from the occasions: baseline (T0. age 13/14y. 1- analysis. From 227 eligible schools (those terials-transfer-disclaimer/physical-activity-downloads/). Participation at T0 was prerequisite for recruitment been associated with higher levels of screen-time (watching to either of the subsequent waves of assessment. age 9/10y. In total. At T0 and T4y. resulting in an additional 62 par- ticipants being assessed by mail. Participants were asked to wear the accelerometer media ownership changes with age and whether changes in for one week and to return it by mail. ported previously [31]. those invited). A minimum of 3 days of valid accelerometer data. Where possible. and fit accelerometers. 157 were approached and A count threshold of <100 counts per minute was used 92 were recruited. 2064 children provided par- nutes of recording between 6 am-11 pm were excluded ental consent and were measured at baseline (57% of [1. the aims of the current study were Norfolk. we ascertained the number pre-schoolers [28] or adolescents [29] or produced mixed of participants attending each secondary school in findings [30].Atkin et al. were mailed an accelerometer and a detailed instruction few longitudinal studies have examined how electronic sheet. At T4y. all participants with changes in sedentary behaviour patterns. and Eating Behaviour: En- Objectively measured sedentary time vironmental Determinants in Young People (SPEEDY) At each assessment. year follow-up (T1y. For quality control pur- ethics committee. Approximately two-thirds of children in the US were requested to return the accelerometer one week and UK have a TV in their bedroom [20. following the same Methods methodology as T1y. the Youth Physical Activity Questionnaire . Periods of ≥10 minutes in school year 5 and their parents (n = 3619) were sent of consecutive zero counts [2. showering and swimming. in Year 9 assemblies at secondary schools attended by ment) (2) examine the cross-sectional association of bed. April-July 2008).cam.

With the mass index (BMI. test-retest reliability ICC = 0. International Journal of Behavioral Nutrition and Physical Activity 2013.007) than composite score (range 0-3) to better reflect the under. Age and sex 319 (T4y. valid accelerometer data was obtained for 1512 Covariates (T0.001).01) and T4y (p < 0. Data on self- were self-reported. Postal code was used to determine of higher SES than those that were excluded (all urban/rural location of participants home [40]. children who were overweight (T0 (p for Statistical analysis χ2 = 0. Time spent playing video were used to assess potential effect modification by sex. T1y (p = 0.Atkin et al. Interaction terms weekdays and weekend days. [38]). The proportion of participants with a TV or computer yes coded 1). Baseline characteristics for those families of low SES (T0 (p < 0. Because different markers of Additionally.9% of participants had a TV in their bedroom at T0. weight status (normal weight/ exception of the sample providing screen-time data at overweight) was determined using age and sex specific T0. Data on objectively mea- lying SES construct. dwellings / villages: rural). accelerometer de- cause the survey made no distinction between ‘active’ and rived sedentary time is presented as the proportion of ‘passive’ video games consoles. weight net. Overall.53) and watching TV (including video/DVD.05). Separate items Whitney tests respectively. we derived a (p = 0. at T1y were more likely to be female (p = 0. 85%) trained research assistants and used to calculate body and 373 (T4y. wear time spent sedentary and weighted hours per week spent sedentary (5*weekday + 2*weekend). wear time spent sedentary. sociation of bedroom media with sedentary time and haviours and the average duration of participation. at T1y (p = 0. Four p < 0.001). 18%) participants respectively. transformed (Box-Cox) prior to analysis. T1y and outcome was change in the proportion of wear time T4y. participants without valid data. Results increase) from T0 to T4y were derived. in bedroom media ownership according to sex. 715 (T1y.026)) or from Station.002)). which is based on the Children’s Leisure and screen-time were examined using t tests and Mann- Activities Study Survey (CLASS. time/T4y wear time) × 100]-[(T0 sedentary time/T0 wear Categorical variables indicating changes in bedroom TV time) × 100]. present. participants included in cross sectional analyses were thresholds [39]. living in an urban area (T0 Analyses were conducted using Stata (Stata. in the bedroom is presented in Figures 1 and 2. participants with valid accelerometer data socioeconomic status (SES) may exhibit differential asso. Items screen-time. In longitudinal analyses. T1y (p = <0.05) and live in a rural location (p = 0. but not change in referred to behaviours occurring outside of school hours screen-time. outcome variables were con- and computer ownership (constantly absent. the outcome was the proportion of The presence of a television or computer in the bed. Sex differences in sedentary time of a computer in the bedroom increased from 21. and house ownership (rental coded 0. declining to (YPAQ. mid (score 2) or high (score 3) SES groups. TX) in 2013. younger ciations with sedentary behaviour [41]. coded 0. (score 0/1). The sample providing valid accelerometer data density profiles were collapsed into a dichotomous vari. T0 absent / T4y present. 1 week Χ2 tests. Participants were assigned to low 70. College (p < 0. sex.5% at T4y. car ownership (no coded 0.93). were non-normally distributed and therefore but did not distinguish recreational from educational use. A combined bedroom media score (range 0-2) was spent sedentary. In cross- Bedroom media sectional analyses. the room was self-reported by participants at T0. 35%). SES and urban/rural location. decrease. Participants indicated on multi-level linear regression was used to examine the as- which of the previous 7 days they had engaged in these be. T0 present / T4y absent) and combined bedroom media items (stable. Relative to their respective ref- erence groups. hamlets and isolated than participants who were excluded from analyses.7% at T0 . the score comprised parent. kg/m2). Within time-point differences assessed time spent using a computer (including the inter. and All covariates were assessed at baseline. calculated as follows: [(T4y sedentary calculated as the sum of the TV and computer items. Estimates of screen-time.001)) were included and lost to follow-up were compared using more likely to have a TV in the bedroom. >16 years coded 1). The presence t tests and Χ2 tests.009). 1 week test-retest reliability ICC for CLASS items = status.01). In both cases. T1y and T4y are presented in Table 1.01) also had lower BMI able (city / town and fringe: urban. T1y (p < 0. the frequency by the duration and summing the total for BMI.ijbnpa. SES and urban/rural location were assessed using 0. 10:137 Page 3 of 10 http://www. games was not included in our screen-time calculation be. Height and weight were measured by reported screen-time was provided by 1745 (T0. In the cross-sectional and longitudinal analyses. For descriptive purposes (Table 1). age. [37]). own/buying coded 1). T4y (p = 0. Regression Weekly hours of screen-time was calculated by multiplying models were adjusted for school-level clustering. constantly tinuous and normally distributed. sured sedentary time and self-reported screen-time at reported age at leaving full time education (≤16 years T0. Of the 2064 children that provided parental consent at baseline. 15%) participants respectively. 73% of baseline participants).001).

y.3 40.9** 64. IQR.3 (0. April-July 2007 (n=1808). (b) weight status.0) Boys 8. standard deviation.3 63.8) .05).2 (0. *Difference in proportions within time-point (p<0.6 Accelerometer Sedentary time. median (IQR) All 6. 14.05). Figure 1 Proportion of children with a TV in their bedroom.3 ± 6.3 ± 5.3% male.3)y 45. April-July 2008 (n=849).9 ± 5.6 35. **Within time-point difference between sexes (P < 0. 14. % of wear time All 62.3 35.1* 36. T4y. T0.3−26. 15. T1y.2 (8.3 (0.0 40.0** 70.9% male.0 (8.31)y 41. International Journal of Behavioral Nutrition and Physical Activity 2013.3 (0.8 T0 (n = 1745) T4y (n = 373) 44.8 ± 5.0% male. SD.3)y Accelerometer Sedentary time.7 62.2 ± 6.8* 40.9 (2.6 ± 5.1 ± 6.2 Girls 63. .Atkin et al.4 69. April-July 2011 (n=438).9−25. T1y. 15.6) .3)y 48.5 Boys 61. inter-quartile range.0% male.4 ± 6. April-July 2008.1 ± 5.6 ± 5.0 ± 4. 10.9−14. Hrs/wk All 34. stratified by (a) sex. Table 1 Children’s objectively measured sedentary time and self-reported screen-time at 3 time-points (values are mean ± SD unless stated otherwise) T0 (n = 1512) T1y (n = 715) T4y (n = 319) 44. *Within time-point difference between sexes (P < 0. T4y.01). (c) urban/rural location.3 Boys 34.2 ± 6.1 (3.0) T0.3 70.3 ± 5. April-July 2011.4 ± 4.5−26. April-July 2007.5 ± 6.8 Girls 35.2 ± 5.3−16. 10. years.1 (2.0% male.1 ± 5. 10:137 Page 4 of 10 http://www. 15.3)y Self-reported screen-time.5) Girls 6. (d) socioeconomic status.6−13. Hrs/wk.2)** .1 (8.3 (0.6 ± 7.

1. T0. there was evidence of a . participants in this group from urban compared to rural locations (p = 0. T1y (p < 0. interaction with sex was observed for self-report out- T4y: Figure 2 Proportion of children with a computer in their bedroom.027). whilst the presence of a computer in the bedroom rose. likely reflecting both the ing a computer in the bedroom and combined bedroom true nature of the association between bedroom media media score was positively associated with self-reported and children’s sedentary behaviours and contemporary screen-time at T0 and T4y.2% to 48.001)). T1y (p < 0.9%.05). having a TV ation of bedroom media with objectively measured in the bedroom and total bedroom media were negatively sedentary time and self-reported screen-time. though few interactions there was a small increase in the proportion of children attained significance (Additional file 1). *Difference in proportions within time-point (p<0. pattern of findings emerged. were typically stronger in girls. bedroom TV ownership declined substantially TV or computer ownership and change in objectively. vealed that cross-sectional associations between bed- the proportion of children with no electronic media in the room media and objectively-measured sedentary time bedroom increased from 25.8%. (b) weight status. April-July 2011 (n=438). At T4y only. April-July 2007 (n=1808).ijbnpa. Tests for interaction by sex re- in the bedroom is presented in Figure 3. Having a computer in the In cross-sectional analyses.009)) and Discussion SES (T0 (p < 0.8% at T4y. T4y (p = 0. no significant developments in electronic media ownership. (c) urban/rural location. T4y.017). Over associations were observed between changes in bedroom 4 years. International Journal of Behavioral Nutrition and Physical Activity 2013.001). T1y. measured sedentary time.2%).Atkin et al. to 30. Concurrently. The reported smaller increases in screen-time than those in proportion of participants with 0. stratified by (a) sex. (d) socioeconomic status. were observed by weight status (T1y p = 0. From T0 to T4y. or 2 electronic media the reference group. Hav. April-July 2008 (n=849). No evidence of with both a TV and computer in the bedroom (T0: 17. Longitudinally. At T0 and T1y. the proportion of participants bedroom at T0 but not at T4y was negatively associated with a computer in their bedroom was higher in children with change in screen-time. 10:137 Page 5 of 10 http://www. A complex associated with objectively measured sedentary time. urban/rural location (T0 (p < 0.001).001). Differences in combined bedroom media score comes or in longitudinal models. Main findings Cross-sectional and longitudinal associations of bedroom This study examined change in the electronic media en- media with sedentary time and screen-time are presented vironment of the bedroom over 4 years and the associ- in Tables 2 and 3 respectively.

01 −0.49) 0.27 T0 (n = 1745) T4y (n = 373) β (95% CI) P β (95% CI) P † Self-reported screen-time Television in bedroom 0.21 (−1.68 (−2.17 (−1. (d) socioeconomic status. Beta coefficient (95% CI). T0.ijbnpa. T4y.20 (0.10. 0. April-July 2011 (n=438).12 Combined bedroom media −0.88.25. 1. International Journal of Behavioral Nutrition and Physical Activity 2013.27) 0.29) 0. stratified by (a) sex.35 0.46) 0. 0. 0. 0.25 (−0. or 2 electronic media (TV/computer) in their bedroom.01 Models adjusted for clustering in schools. Table 2 Cross-sectional association of bedroom media with objectively measured sedentary time and self-reported screen-time T0 (n = 1512) T1y (n = 715) T4y (n = 319) β (95% CI) P β (95% CI) P β (95% CI) P Accelerometer sedentary time* Television in bedroom −1.06 (−0. Beta coefficient (95% CI)*100. 0.07) 0.51. (c) urban / rural location.03 0.79 (−1.35 (0. −0.01 Combined bedroom media 0.07 Computer in bedroom 0.21 (0. sex. . † Outcome variable: Weekly hours of self-reported screen-time (Box-Cox transformed).18) 0.70) <0.02.Atkin et al. 1.22 (−0. *Difference in proportions within time-point (p<0. *Outcome variable: Weekly proportion of wear time spent sedentary. (b) weight status. −0.60) <0.01.82) Figure 3 Proportion of children with 0.05).02 <0.07.34) <0. April-July 2008 (n=849). age. T1y. socio-economic status and urban/rural location.01 −1.42) 0.65) 0. −0.46) <0. 1. 2.29 (−0.03 0. April-July 2007 (n=1808).51 (−0. −0.77 Computer in bedroom −0.15 (0. body mass index.19) 0.02. 10:137 Page 6 of 10 http://www.49 (−1.01 0.05. 1.76 (−1. 0.64 1.32. 0.

1.75 32 1. † Outcome variable: Change in weekly hours of self-reported screen-time T4y-T0.53) 0.67 45 −8.54) 0.25. 2011 survey.57) 0.49 (−2. tary time from T0 to T4y is consistent with previous re.51.ijbnpa.18. socio-economic status and urban/rural location.49 (−1.17 (0.88 (−5.28 T0 present – T4y absent 38 0. that trends and examine its impact upon behaviour and includes assessment of specific behaviours as well as health. though the magnitude of were not associated with changes in overall sedentary change was smaller than that seen in the current study time. 3. sex.33 Increase 77 0. 6.82) 0. e-book readers and mobile phones were listed above search indicating that sedentary behaviour increases as television sets as items that US children and adolescents children age [1. remains unclear.91) 0.07) 0. Findings may reflect age related changes in media preferences [21] but also broader societal patterns in Comparison with other evidence and implications electronic media ownership over the study period.43) 0.96 (−3. 2. computers (including ‘tablet’ computers).43 (−2. Reported screen-time in the current had most interest in buying within the next 6-months study was somewhat lower than has been reported previ.97.64 (−0. will enable interventions to be multitude of platforms through which children may ac- developed and targeted more precisely.01) 0. Going forward.82 (−3.20 T0 present – T4y absent 71 0. whilst the proportion of children with a computer in the bedroom differed by socioeconomic status and by . Recent data from Ofcom (the in- tary time were less clear.38. 5. both portable and home-based. reference group.45]. 1. 3. in the bedroom increased by approximately 10% points dren’s screen-time but associations with overall seden.15) 0. Beta coefficient (95% CI).68 38 3.08 (−1.18 Ref. 3. cess audio-visual content and acknowledge the increas- The proportion of children with a TV in the bedroom ing portability and multifunctionality of new devices.61 94 0. 5. International Journal of Behavioral Nutrition and Physical Activity 2013.04 T0 absent – T4y present 36 −0. 1. games consoles) to attributable to increased screen-time or shifts in other access video content [44.39. a rise in the availability of the internet in children’s bed- time. this may be due to temporal shifts ing live or time shifted content on a television set deliv- in screen-time habits.59 (− Table 3 Longitudinal association of change in bedroom media with change in sedentary time and self-reported screen- time Sedentary time (n = 283)* Screen-time (n = 357)† Variable n β (95% CI) P n β (95% CI) P Television Constantly absent 90 Ref 112 Ref Constantly present 86 1.13 84 2. Models adjusted for clustering in schools. [43]. [21].02 (−12. More detailed characterisa.70 Computer Constantly absent 164 Ref 196 Ref Constantly present 23 0. 10:137 Page 7 of 10 http://www. computers. 2. 8. −3. Beta coefficient (95% CI)*100.53 (−2.29) <0. as consumers increasingly utilise The extent to which the rise in overall sedentary time is alternative devices (e. *Outcome variable: Change in weekly proportion of wear time spent sedentary T4y-T0.89.48.48 T0 absent – T4y present 58 1.67 102 −1. The evolution and uptake behaviours. 7.2. but changes in the bedroom media environment rooms between 2007–2010.56.87) 0.75.21 113 4.68. over the same period.79.01 Combined bedroom media Stable 126 Ref 155 Ref Decrease 80 −0. quently outpaces the efforts of researchers to document tion of age-related changes in sedentary behaviour.57 (−1. fre- or car use.49) 0.Atkin et al.62.35) 0.63 100 2.51. Longitudinal analysis indicated dependent regulator for the UK communications indus- that removal of a computer from the bedroom may be a tries) also indicates a decline in the presence of TVs and means of limiting the age-related increase in screen.g. 2.83) 0. Television viewing in the traditional sense (watch- ously in UK children. decreased by approximately 30% points between T0 and The proportion of participants with a TV or computer T4y.75) 0. researchers should recognise the overall sedentary time. for example greater homework requirements of electronic media.1) 0. age.39 (−0. positive association between bedroom media and chil.57 (−3.59 (−1. 9. In a The observed increase in screen-time and overall seden.64. contrasting sample characteristics ered by broadcast signal or paid TV subscription) or the application of different measurement tools [42] appears to be declining. body mass index.42].

The use of different markers from children’s bedrooms as a means of limiting screen- to indicate SES. enabled us to identify changes in behaviour and the bed- trary (associations were non-significant at T1y and T4y).49]. [9. or reflects a changing pattern were adjusted for a number of potentially confounding of influence as children age. Repeated assessments on a single cohort parent switch in the direction of the association is arbi. International Journal of Behavioral Nutrition and Physical Activity 2013. Given the location of a particular device [26]. They are also portable and multifunctional. An urban/rural divide in internet usage and con. onment were attributable to age-related changes in tronic media ownership and avoid an overly restrictive media preferences or broader societal changes in elec- focus solely upon TV [27]. further research on this ques. more explicit in gathering data on the contexts in which room and accelerometer-determined sedentary time and behaviours occur.ijbnpa. finding that the presence of a computer in the bedroom In the current study. [47. In most unable to differentiate between active and passive games models. our assessment of the bedroom media environment or time was observed in participants that reported having a in our screen-time outcome.43]. having a TV in the bedroom and com. It is unclear whether this ap. new instruments tronic media use over the study period. Previous studies have reported that in the anticipated direction but did not attain statistical having a TV in the bedroom is typically more common urban/rural location. Strengths and limitations tion is likely and may provide some clarity. we did not media and change in sedentary time or screen-time were assess ownership of tablet computers or distinguish . 10:137 Page 8 of 10 http://www. with objectively measured sedentary time at T0 and T1y. tation is that video game consoles were not included in In longitudinal analyses. as it may no longer be appropriate to infer by adolescents and are an increasingly important means posture on the basis of reported behaviour. which was individually associated with clude process evaluations to understand the acceptability screen-time at T0 and T4y. Regression models an artefact of study design. efficacy of this approach is lacking.21]. Interestingly. and be shown no association between having a TV in the bed. Researchers of bedroom media with sedentary time was stronger in need to recognise the broad range of electronic media girls than boys but interaction tests were mostly border. room media environment over time.48] but we are unaware of any existing studies that However. limits comparability between time. we are not able to ascertain increased from T0 to T4y. Trials exploring the influence of removing be driven predominantly by the presence of a computer electronic media from children’s bedrooms should in- in the bedroom. In this study. It may also be valuable to grown rapidly in recent years [20. these data suggest that re. Previous studies have predominantly users may perform multiple tasks simultaneously. ple of children. rather than inferring context from the no evidence for interaction with sex [28. the associations between a change in bedroom consoles in baseline assessments. Findings of the nection speeds has been reported previously in the UK current study provide some support for this strategy. This is because we were computer in the bedroom at T0 but not at T4y. Numerous internet access in the bedroom is more prevalent in studies have advocated the removal of electronic media higher SES groups [21. whether observed changes in the bedroom media envir- searchers must acknowledge emerging trends in elec. however. whilst having a computer or reduced statistical power for these analyses. using the internet. watching TV) simultaneously and There was some evidence that the negative association without being tied to a specific location. Uptake of infor- may be required to capture the diversity of electronic mation technology and communications media has media used by young people. Considered alongside our of this strategy for children and parents.30]. including the extent to which line significant. from participants’ bedrooms is lacking in our analysis.Atkin et al. combined bedroom media score was Moreover.21. In addition. In so doing. A limi- this potential interaction with age are required.25-27] but experimental research exploring the studies.g. and parents may be resistant to the idea of removing consistent with previous research [18.46]. It is also of The strengths of this study include its longitudinal de- interest that cross-sectional analyses at T4y identified sign and the collection of objective and self-report data (non-significant) positive associations between bedroom on sedentary behaviour in a large population-based sam- media and sedentary time. important contextual information regarding have examined urban/rural differences in the bedroom the circumstances in which computers were removed media environment. qualitative work has indicated that children positively associated with screen-time at T0 and T4y. electronic media from the bedroom once they have been the positive association observed in this study appears to installed [50]. bined bedroom media score were negatively associated enabling the user to perform multiple tasks (e. used by young people. [20. Further studies exploring factors and interactions with sex were explored. by which young people accumulate screen-time Unexpectedly. increasing use of accelerometry to assess sedentary time in epidemiological studies. gaming. Devices such as concurrently ascertain context of behaviour and postural tablet computers and mobile phones are highly valued allocation. Participant attrition may have resulted in amongst low SES families. a smaller increase in screen.

LeBlanc AG. Int J Behav and communications technology by young people. Marcos A. 8:132. 126:e926–e935. Nutr Metab and critically reviewed the manuscript. Olds T. BMI and urban/rural location of the Clinical Medicine. Start Active Stay Active: A Report on Physical Activity for Health from acknowledged. Luan J. and parents for their participation. thank the schools. Riddoch Additional file C. Griffin SJ: Changes in children’s physical activity over 12 months: longitudinal results from the media was more consistently associated with children’s SPEEDY study.ijbnpa. Appl Physiol Nutr Metab 2011. and the Wellcome Trust. Cooper A: Moderate to with objectively measured sedentary time. data acquisition and analysis. In addition. Where appropriate. van Sluijs EM. of Excellence (RES-590-28-0002). Murumets K. Beghin L. Reilly JJ: In this study we found that there were notable changes Longitudinal study of physical activity and sedentary behavior in in the electronic media environment of children’s bed. De ther research examining the context and content of elec. . Chinapaw MJ. iour and obesity development in children and adolescents. 9. Duggan M: Canadian sedentary behaviour guidelines for National Institute for Health Research. Foley LS. everyone who come data at T1y and T4y. 1999. UK. Unit. rooms over 4 years. In all cases. Cardiovasc Dis 2008. under the children and youth. MRC Epidemiology the direction and magnitude of associations were min. 14. bias by limiting the analytical sample to participants that Author details provided complete data at all time points. Martinez-Gomez D. we thank Kate Westgate and Stefanie Mayle findings. Acknowledgements 10. the Colley RC. Rey-Lopez JP. stratified by between desktop or laptop computers in our question. Health and CA: Sage. Saunders TJ. The SPEEDY study is a ences in accelerometer wear between waves of collaboration between the MRC Epidemiology Unit and the School of assessment by deriving outcomes that were relative to Environmental Sciences and the School of Medicine. Larouche R. 174:173–184. Connor Gorber S: Systematic review of sedentary supported. JAMA 2012. Pediatrics 2011. Sherar LB. Esliger DW. 4. Medical Research Council. 36:59–64. correlates. Obes Rev 2011. 18:242–251. auspices of the UK Clinical Research Collaboration. Corder K. Kho ME. 5:518–530. 307:704–712. between young peoples’ sedentary behaviour and biomedical health indicators: a systematic review of prospective studies. Andersen LB: TV viewing and physical activity are independently associated with metabolic risk in children: the European youth heart study. and All authors read and approved the final manuscript. Am J Epidemiol 2011. Norwich. Sardinha LB. possibly Received: 26 July 2013 Accepted: 13 December 2013 Published: 17 December 2013 resulting in misclassification. Pediatrics 2010. Tremblay MS. Am J Lifestyle Med 2011. 1 UKCRC Centre for Diet and Activity Research (CEDAR). Funding from the British Heart Foundation. Box 285. Jones AP. Ortega FB. Authors’ contributions 12:e621–e632. 11. a UK Clinical Research Collaboration Public Health Research Centre Int J Behav Nutr Phys Act 2011. Brug J. KC contributed to study design. Proper KI. Economic and Social Research Council. Additional file 1: Cross-sectional association of bedroom media 6. Griew P. Labayen I. Moreno LA: Sedentary behav- manuscript. Ekelund U. 8:98. performed data analysis and drafted the 8. Tremblay MS. and that the presence of bedroom 2. children. and critically reviewed the manuscript. LeBlanc AG. Sallis JF. MC_UU_12015/3. Parkinson KN. 2MRC Epidemiology Unit. information on Metabolic Science. Goldfield G. Ekelund U. Maddison R. Mauro B. Biosca M. MC_UU_12015/4) and the National Prevention Research Initiative. Cambridge CB2 0QQ. We addressed the potential limitation of differ. 2011. Welsh Assembly Government. Tremblay MS. EMFS contributed to study design. UK. Conclusion References 1. Am J Prev Med 2011. Thousand Oaks. 127:e24–e30. Ruiz JR. Nutr Phys Act 2012. Competing interests 7. Colley The work of Andrew J Atkin. Molnar D. measured physical activity and sedentary time in European adoles- cents: the HELENA study. van Mechelen W. Marshall SJ. Social Care Research and Development Office for Northern Ireland. Brage S. consisting 13. University of Cambridge School of Clinical Medicine. Bourdeaudhuij I. Harro M. Owen N: Physical Activity and Behavioural Medicine. Ramirez E: Reducing sedentary behavior: a new paradigm in of the following funding partners: British Heart Foundation. data collection and data processing is gratefully acknowledged. data acquisition and analysis. Froberg K. 10:137 Page 9 of 10 http://www. Chief naire. Anderssen SA. The Department of Council (Unit Programme number MC_UU_12015/7. Kho ME. Economic and Social Research Council. Cancer Research physical activity promotion. Hume C: Health risks. In 3. Health. Janssen I. 5. PLoS Med 2006. Salmon J. fur. Marshall SJ. International Journal of Behavioral Nutrition and Physical Activity 2013. Scottish Government Health Directorates. and Esther M F van Sluijs was RC. Institute of imally affected (data not shown). Gonzalez-Gross M. UK. Jiang Y. Ridley K: It’s not just the television: survey light of the rapid and continuing uptake of information analysis of sedentary behaviour in New Zealand young people. Lastly. Basterfield L. The SPEEDY study was funded by the Medical Research the Four Home countries’ Chief Medical Officers. Adamson AJ. The response Practice at the University of East Anglia. Scientist Office. Department of Health. children. Frary JK. Sjostrom M: Objectively health. (CEDAR). AJA conceived the analysis. Singh AS: Relationship The authors declare that they have no competing interests. Cambridge Biomedical Campus. Medical Research Council. Manios Y. Cambridge. interventions to reduce sedentary behavior in young people. Kirsten Corder. United Kingdom. Vicente-Rodriguez G. Pearce MS. Diabetes United Kingdom. We from lower SES families were less likely to provide out. 41:197–206. by the Centre for Diet and Activity Research behaviour and health indicators in school-aged children and youth. 3:e488. limiting generalizability of helped with the data collection and Norfolk Children’s Services for invaluable input and support. is essential. we conducted sensitivity from the physical activity technical team at the MRC Epidemiology Unit for analyses to examine the potential impact of selection their assistance in processing the accelerometer data. Health Policy and recorded wear time at each time point. is gratefully 12. Moreno LA. it is possible that these factors may have changed over time. and World Cancer Research Fund. vigorous physical activity and sedentary time and cardiometabolic risk factors in children and adolescents. home was collected at baseline only.Atkin et al. Hicks A. and its impact upon behaviour and Widhalm K. Ekelund U. Castillo MJ. University of Cambridge School of the covariates SES. tronic media use. wholly or in part. screen-time behaviour than overall sedentary time. The contribution of our collaborators at the University of East Anglia in obtaining funding and ethical rate at T4y was low and participants with higher BMI or approval.

Arundell L. 45. CAS. binaries/research/media-literacy/media-lit11/childrens. Panter J. Craig R. project. Schmidt ME.nielsen. Treuth MS.: Bedroom media. Ondrak KS. Cole TJ. social. Catellier DJ. Int J Behav Nutr Phys Act 2006. Evenson KR. van der London. Nielsen Wire: US kids looking forward to “iHoliday”.ofcom. Skidmore PML. Horst K. Cooper AR. Volume 1: Year Olds. Lewis among young children. J Phys Act Health 2011. Rimer BK. In Heal Behav Heal 37. Andersen LB. 2004. Griffin SJ: Behavioural and social screen-time in children: a longitudinal analysis.pdf. Crawford D. relationship between maternal education and children’s TV viewing. Leeds: The NHS Information Centre for health 21. Lytle LA: Adolescent physical activity and screen time: associations with the physical home environment. 9:15. Neumark-Sztainer D. Adachi-Mejia AM. Bellizzi MC. 45:906–913. 31. Farbakhsh K.1186/1479-5868-10-137 Cite this article as: Atkin et 10-year-old Portuguese children. 27. among children: sedentary behavior at home. Loprinzi PD. 20. Sherry B. England: University of 19. Wareham NJ. doi:10. Page AS. Office for National Statistics: Regional trends. 40. Jolley D. Ekelund U. 33. Sardinha LB. 33:41–47. Bibby P.US. CA: Jossey-Bass. Trost SG. O’Brien A. Flegal KM. Pediatrics 2008. Uijtdewilligen L. J Sport Sci 2008. possible to assess free-living physical activity and energy expenditure in San Francisco. Anderssen SA.rural and urban ourselves: relationships between screen media and childhood areas: comparing lives using rural/urban classifications. and summary and recommendations from an expert panel meeting. Edited by Glanz K. Med Sci Sport Exerc 2004. 43. Jordan AB. 3:36. Pfeiffer KA: Comparison of accelerometer cut points for predicting activity intensity in youth. Salmon J. http://www.quarter 1. 36:1259– Chinapaw MJ: Determinants of physical activity and sedentary 41. dear viewer. Price S. Harvey I. 121:718–724. 16. 8:A131. Med Sci Sports Exerc • Convenient online submission 2011. Marshall SJ: Family circumstance. http://stakeholders. Foehr UG. Broyles ST. van Mechelen W. Moore R. • Immediate publication on acceptance 36. Ridley K. http://maps. Harrison F. 2002:462–485. Murray DM. 320:1240–1243. Salmon J. Sirard JR. Timperio A: Individual. Ostbye T: Correlates of 50.ons. Rich M: The fault. Wareham 15. Page A. Salmon J: Children’s Television viewing and 39. overweight. McDonald J. Ekelund U: Objectively measured time spent sedentary is associated with insulin resistance independent of overall and central • Research which is freely available for redistribution body fat in 9. 2010. Griffin SJ: Physical activity and dietary behaviour in a population-based sample of British 10-year old children: the SPEEDY study (Sport. 8:613– . Int J Behav Nutr Phys Act 2009. Shepherd J: Developing a New Classification of Urban and Rural Br J Sports Med 2011. Submit your manuscript at www. Beach ML. 10:105. Byun W. 44:747–755. Areas for Policy Purposes: The Methods. Timerio A: Agreement between activPAL and Actigraph for assessing children’s • No space constraints or color figure charges sedentary time. Int J Behav Nutr Phys Act 2012. 17. Pediatr Clin North Am 2011. and weight status: 49.html. 26. Gibson JJ. Ann Am physical environment factors associated with electronic media use Acad Pol Soc Sci 2008. Titus-Ernstoff LT. Callaghan AM. Ornelas R. O’Connell E. 28. overweight/obesity.quarter 4. sedentary time and 30. http://www. van den Berg P. and take full advantage of: 34. Hirani V (Eds): Health Survey for England 2008. Mitchell JA. http://kff. http://www. young people by self-report? Am J Clin Nutr 2009. Jones AP. activities study survey (CLASS). Campbell K. Taveras EM: bedrooms. Res Pract. Sisson SB.Atkin et al. Wright A. Physical Activity and Fitness. Van Sluijs EM. Physical activity and Eat- ing behaviour: environmental Determinants in Young people). McKetta S. Am J Prev Med Pediatr Exerc Sci 2004. 32. circumstance. Diabetes Care 2008. Rideout VJ. Int J Behav Nutr Phys Act 2013. Submit your next manuscript to BioMed Central Going S. Catellier DJ. Gorely T. International Journal of Behavioral Nutrition and Physical Activity 2013. 45:896–905. 89:862–870. 118:e1303–e1310. • Thorough peer review 35. 18. Dietz WH: Establishing a standard definition objectively measured physical activity: associations with family for child overweight and obesity worldwide: international survey. Robinson TN: Children. Ofcom: Communications infrastructure report.html. Dolinsky DH. Gill K. London. sedentary 38. 2009. gov. television viewing. lies not in the screens. bedrooms of children: does it impact health and behavior? Prev Med 46. Int J Behav Nutr Phys Act 2010. McDivitt JA. 31:644–651. Br J Sports Med 2011. Baker BL. Salmon J: Mediators of the 2012-Reports/nielsen-cross-platform-q4-2011. Scopus and Google Scholar Riddoch CJ. Longacre MR. Roberts DF: Generation M2: Media in the Lives of 8–18 42. 10:137 Page 10 of 10 http://www. Telford A. Systematic review of effective strategies for reducing screen time 47. Hesketh K. Norman JE. Ofcom: UK children’s media literacy.US. Pate R: Defining accelerometer thresholds for activity intensities in adolescent girls. Jordan AB. 58:1493–1508. Hersey JC.pdf. Volume 3rd. Pulsford RM. 16:64–78. Mindell J. Haines J. istics associated with older adolescents who have a television in their 25. Whincup P. broadband/downloads/ofcom-uk-broadband-speed-report-2011. 7:82. Hillsdon MM: Socioeconomic behaviour in young people: a review and quality synthesis of position and childhood sedentary time: evidence from the PEACH prospective studies.nielsen. Hesketh K. The Nielsen Company: The cross-platform report . 31:569–575. Int J Obes (Lond) Crawford D. 52:104–108. Van Sluijs EMF. Brouwer RJ. 6:33. Barr-Anderson DJ. 26:1557–1565. Biddle SJ. Br J Sport Med 2010. Siega-Riz AM.nielsen. Obesity 2012. Dalton MA: Children with a TV in their bedroom at higher risk for being 44. Rosenberg M. BMC Public Health http://www. Patnode CD.ijbnpa. Granich J. Behavioral Nutrition and Physical Activity 2013 10:137. Singh AS.10. Kaiser Family Foundation. van Sluijs EM. com/us/en/newswire/2011/us-kids-looking-forward-to-iholiday-2011. McMurray RG: Calibration of two objective measures of physical activity for children. Chernausek SD: TVs in the 2012-Reports/Nielsen-Cross-Platform-Report-Q1-2012-final. Ball K. Corder K. Sallis J. 22. 615:119–132. Ridgers ND.pdf. Newton RL Jr. 43:1360–1368. International Journal of correlates of sedentary time in young people.biomedcentral. Owen N: Ecological Models of Health and social care. 8:388. Prev Chronic Dis 2011. BMJ 2000. Mwanza K.ofcom. Cassidy A. Dowda M: Sedentary behaviour in youth. 2011 . McMurray RG. Pediatrics 2006. Nauta J. Atkin AJ. Quiterio AL. Story M: Character- 2011. Laska MN. Ommundsen Y. doi: 10. Griew 23. Pate RR. Ekelund U: Is it Educ Theory. Schmitz K. Froberg K. The Nielsen Company: The cross-platform report . Heitzler CD: Reducing children’s sedentary time and physical activity among preschool-aged children. Evenson KR. television-viewing time: a qualitative study of parents and their children. 29. 2012 . • Inclusion in PubMed. but in 48. Almeida MJ. Knuiman MW. Crawford D: Reliability and validity of behaviour and physical activity in adolescents living in England: project physical activity questionnaires for children: the children’s leisure STIL. No 43 . Twisk JW.