UBC Faculty Research and Publications

Temporal trends in and relationships between screen time, physical activity, overweight and obesity Duncan, Mitch J; Vandelanotte, Corneel; Caperchione, Cristina; Hanley, Christine; Mummery, W K Dec 8, 2012

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12889_2012_Article_4810.pdf [ 198.26kB ]
JSON: 52383-1.0223943.json
JSON-LD: 52383-1.0223943-ld.json
RDF/XML (Pretty): 52383-1.0223943-rdf.xml
RDF/JSON: 52383-1.0223943-rdf.json
Turtle: 52383-1.0223943-turtle.txt
N-Triples: 52383-1.0223943-rdf-ntriples.txt
Original Record: 52383-1.0223943-source.json
Full Text

Full Text

RESEARCH ARTICLE Open AccessTemporal trends in and relationships betweenscreen time, physical activity, overweight andobesityMitch J Duncan1*, Corneel Vandelanotte1, Cristina Caperchione2, Christine Hanley1 and W Kerry Mummery3AbstractBackground: The aims of this study were to examine temporal trends in the prevalence of sufficientmoderate-to-vigorous intensity physical activity (MVPA), high levels of screen time, combined measures of thesebehaviors and overweight or obesity in Australian adults during the period 2002–2008. Trends over this time periodin overweight or obesity within each behavior group (sufficient/insufficient MVPA, high/low screen time andcombined behaviors) were also examined.Methods: Data were collected via annually conducted cross-sectional computer-assisted-telephone-interviews(CATI) of adults (n=7908) living in Central Queensland, Australia (2002–2008). Self-reported MVPA, screen time (TVviewing and computer use), and BMI were used to create dichotomous classifications of physical activity (SufficientMVPA (S-MVPA), Insufficient Physical Activity (I-MVPA)), screen time (High Screen Time (HST), Low Screen Time(LST)), combined behavior categories (S-MVPA/LST, I-MVPA/LST, S-MVPA/HST, I-MVPA/HST) and BMI (Overweight orObese, Healthy Weight) respectively.Results: The prevalence of S-MVPA, HST, and overweight or obesity increased at approximately the same rate overthe study period in the overall sample and females (p≤0.05). In the overall sample and in females, the prevalence ofoverweight and obesity increased over the study period in those individuals classified as I-MVPA/HST (p≤0.05).Conclusion: Results provide evidence that while the prevalence of S-MVPA appears to be modestly increasing, theproportion of the population engaging in HST and classified as overweight or obese are increasing atapproximately the same rate. These observations highlight the need to increase levels of total physical activity(including light intensity physical activity) and decrease sedentary behavior including screen time.Keywords: Temporal trends, Obesity, Screen time, Physical activity, AdultsBackgroundThe health benefits of physical activity are numerousand well documented particularly in relation to indivi-duals achieving sufficient levels of moderate-to-vigorousintensity physical activity (MVPA), or sufficient MVPA[1]. Sufficient MVPA is frequently defined as achieving150 minutes of MVPA in 5 or more sessions in the pre-vious week [2]. Though MVPA levels are generally lowin the US, Australia and Canada, the population preva-lence of individuals engaging in sufficient MVPA hasgradually increased over recent years [3-6], and has oc-curred during periods when the prevalence of over-weight or obesity has also increased [3,7-9]. Yet trendsin the prevalence of sufficient MVPA and overweight orobesity are not equal across population groups, with sig-nificant differences observed by gender and age [5,9,10].This apparent paradox in the trends of sufficient MVPAand overweight or obesity may due to reductions inlevels of total physical activity due to decreased lightintensity physical activity and increased sedentarybehavior, increased energy intake or a combination ofthese behaviors [7,11].Measures of screen based activities performed inseated postures, such as TV viewing and computer use,* Correspondence: m.duncan@cqu.edu.au1CQUniversity, Institute for Health and Social Science Research, Centre forPhysical Activity Studies, Rockhampton, Bld 18, CQUniversity Australia,Rockhampton, QLD 4702, AustraliaFull list of author information is available at the end of the article© 2012 Duncan et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.Duncan et al. BMC Public Health 2012, 12:1060http://www.biomedcentral.com/1471-2458/12/1060are commonly used indicators of sedentary behavior asthe low metabolic cost of these activities is below the en-ergy expenditure threshold used to define sedentary be-havior (<1.5 METS) [12,13]. Time spent watching TVand using computers are screen based activities whichare associated with increased risk of weight gain, over-weight and obesity, diabetes, and CVD mortality [14-16].Furthermore, the associations between increased screenbased activity and the negative health outcomes remainpresent even when adjusted for engagement in MVPA.As such, the time spent engaged in sedentary behavior isemerging as a potentially important and independentrisk factor for ill health [17,18].There is little data on temporal trends in sedentarybehaviors when compared to data on trends in MVPA.In occupational settings, physical activity has declinedresulting in increased sedentary behavior in workplacesover recent decades [3,10,19]. In non-occupational set-tings sedentary behavior also appears to be increasing. Inthe US, time spent watching TV increased approximately36 minutes every decade in the period 1950–2000 [3];in Australia, average time spent in non-occupational sed-entary behavior increased from 894 minutes/2 days in1997 to 906 minutes/2 days in 2006 [20]. Furthermorecomparisons in the trends of sufficient MVPA and en-gagement in sedentary behavior are infrequently per-formed within the same population [20], or the observedtrends are difficult to interpret due to changes in surveyinstruments and methodologies [6,10]. Examiningchanges in both behaviors is useful in understanding howthese two behaviors have changed over time in relationto disease outcomes and risk factors such as overweightor obesity.Therefore the purpose of this study is threefold, firstly,to examine the relationships between screen time,MVPA and combined measures of these behaviors andbeing classified as overweight or obese in a pooled sam-ple of Australian adults who completed separate crosssectional surveys in the period 2002–2008. Similar asso-ciations have been conducted previously [16,21], and areconducted in the current study to establish the associa-tions between activity behaviors and overweight or obes-ity in the current sample prior to progressing to theother aims of the study. Secondly, the study aimed toexamine temporal trends in the prevalence of thesebehaviors and being classified as overweight or obeseduring this period. Thirdly, the study sought to examinetemporal trends in the prevalence of being classified asoverweight or obese during this period within each ofthe behavior groups examined (sufficient/insufficientMVPA, high/low screen time and combined behaviors).These relationships will also be examined for the overallstudy population and within the male and female sam-ples separately.MethodsSampleThis study uses data pooled from a series of separatecross-sectional surveys of the adult population living inCentral Queensland, Australia. The surveys were con-ducted annually by the Population Research Laboratory,at CQUniversity during the period 2002 to 2008. Thesamples sizes for each survey year are provided inTable 1, the overall sample size was 7908. Details on thesurvey methods are provided elsewhere and each surveywas approved by CQUniversity’s Human Research EthicsCommittee [5]. Briefly, each survey was performed dur-ing October and November using computer-assisted-telephone-interviewing (CATI), all respondents wereaged 18 years or older (range 18–93 years), and were liv-ing in a dwelling contactable by direct-dialed land basedtelephone. Participants were randomly selected using atwo-stage stratified sampling process where households(as phone numbers are linked to households) wereselected first and then individuals within householdswere selected. Potential telephone numbers were drawnfrom commercially available Electronic White Pages, allduplicate, cellular and business numbers were removedfrom the sample before each survey commenced. Eachyear the sample was stratified by gender to reflect thecharacteristics of the Australian population based on theCensus closest to the survey date, however the sample isnot intended to be a representative sample of the Aus-tralian population. The response rate for the surveysranged from 39.3% in 2007 to 62.3% in 2005 with anaverage response rate across surveys of 46.9%. Responserates are comparable to those reported in other recentlyconducted CATI based surveys [22,23], and no informa-tion is available on those who did not participate in thesurvey.MeasuresAll surveys included items that assessed socio-demographic details of respondents including, gender,age, household income, employment status, educationand smoking status. Self-reported height and bodyweight were used to determine BMI and classify partici-pants into healthy weight (BMI ≤24.9) and overweight orobese (BMI ≥25.0) categories. Using the Active AustraliaQuestionnaire all participants were asked to report thefrequency and duration physical activity during recre-ational and transport walking, moderate and vigorous in-tensity physical activity in the previous week [24]. Thisinstrument has demonstrated acceptable test-retest reli-ability and validity [25,26]. Engagement in sufficientMVPA (S-MVPA) was classified as achieving a mini-mum of 150 minutes of MVPA in at least five sessionsin the previous week, participants not satisfying this cri-terion were classified as insufficiently physically activeDuncan et al. BMC Public Health 2012, 12:1060 Page 2 of 9http://www.biomedcentral.com/1471-2458/12/1060(I-MVPA). This method of classification is in accordancewith previously described methods to reflect compliancewith National Physical Activity Guidelines for AustralianAdults [2,25,27]. As a marker of leisure time sedentarybehavior, TV viewing is more consistently associatedwith overweight and obesity in females than in males[28-31]. The differences in associations between gendersmay be because TV viewing is not representative ofother leisure time sedentary behaviors in males [32].Thus, the current study used two measures of screenbased activities as indicators of sedentary behavior, TVviewing and computer use, to better represent broadersedentary behaviors of both males and females. Durationof screen based activity reported in hours and minuteswas assessed using two separate items, “What do youestimate was the total time that you spent watching TVin the last week?” and “What do you estimate was thetotal time you spent working in front of a computerscreen in the last week?” Data from these two items wassummed to provide an overall measure of screen time inthe previous week and dichotomized at 21 hours intohigh screen time (HST) and low screen time (LST) [33].This classification was selected a priori as it approxi-mates an apparent threshold of screen based activity thatis associated with greater risks of ill health compared tolower volumes [14,34]. No psychometric data is availableon the items used to assess screen time. Similar to previ-ous research [21], participants were further classifiedinto four mutually exclusive groups to facilitate analysesthat combines high and low levels for both physicalTable 1 Unweighted sample proportions of central queensland social survey participants by selected socio-demographic and behavioral categories 2002-20082002(n=1127)2003(n=1147)2004(n=1102)2005(n=1127)2006(n=1131)2007(n=1112)2008(n=1162)Overall Sample(n=7908)GenderMale 52.1 51.9 51.1 51.6 51.5 52.0 50.9 51.6Female 47.9 48.1 48.9 48.4 48.5 48.0 49.1 48.4Age18-44 50.8 45.5 44.7 45.6 41.6 39.3 38.8 43.845+ 49.2 54.5 55.3 54.4 58.4 60.7 61.2 56.2Employment StatusYes 62.3 62.8 63.7 66.5 65.8 64.6 62.8 64.1No 37.7 37.2 36.3 33.5 34.2 35.4 37.2 35.9Years of Education0-12 Years 59.1 57.3 59.6 57.9 57.1 61.2 55.5 58.213+ Years 40.9 42.7 40.4 42.1 42.9 38.8 44.5 41.8Smoking StatusYes 20.9 22.7 19.5 21.9 20.1 17.2 17.2 19.9No 79.1 77.3 80.5 78.1 79.9 82.8 82.8 80.1BMIHealthy Weight 42.1 43.8 44.6 40.5 37.7 37.7 37.7 40.6Overweight or Obese 57.9 56.2 55.4 59.5 62.3 62.3 62.3 59.4Physical ActivityInsufficient Physical Activity(I-MVPA)56.9 53.4 54.4 53.5 51.6 60.3 46.8 53.8Sufficient Physical Activity(S-MVPA)43.1 46.6 45.6 46.5 48.4 39.7 53.2 46.2Screen TimeLow Screen Time (LST) 59.2 58.5 58.8 54.7 52.4 55.1 54.0 56.0High Screen Time (HST) 40.8 41.5 42.2 45.3 47.6 44.9 46.0 44.0Combined BehaviorsS-MVPA/LST 25.7 27.6 27.0 26.7 24.9 21.5 29.4 26.1I-MVPA/LST 33.5 31.0 30.9 28.0 27.5 33.6 24.6 29.8S-MVPA/HST 17.4 19.0 18.6 19.8 23.4 18.2 23.8 20.0I-MVPA/HST 23.4 22.5 23.6 25.5 24.1 26.7 22.2 24.0Duncan et al. BMC Public Health 2012, 12:1060 Page 3 of 9http://www.biomedcentral.com/1471-2458/12/1060activity and screen time: S-MVPA/LST, I-MVPA/LST,S-MVPA/HST, I-MVPA/HST (these variables are alsoreferred to as ‘combined behaviors’).Prior to analysis any duplicate telephone numbers be-tween surveys were excluded. Analysis is delimited tothose with complete data for all outcome variables, in-cluding screen time, resulting in a different sample sizecompared to previous analysis of this dataset [5]. Usingdata pooled from seven cross-sectional surveys (2002–2008) three separate binary logistic regression modelswere used to assess associations between the three dif-ferent behaviors (MVPA, screen time, combined beha-viors) and the likelihood to be classified as overweight orobese (Aim 1). To examine temporal trends in theprevalence of overweight or obesity, S-MVPA, and HSTover the study period, separate binary logistic regressionmodels were conducted including the ordinal variable,year of survey, as a continuous predictor. Multinomiallogistic regression was used to model temporal trendsin combined behaviors over the study period usingS-MVPA/LST as the reference category and includingyear of survey as an ordinal variable. This allowed trendsover the study period to be examined rather than con-trasts between specific years in the study period (Aim 2).For the final analyses, the sample was stratified by eachbehavior group and separate binary logistic regressionmodels were used to model the trend (year of survey) inoverweight or obesity over the study period (Aim 3). Allanalyses were adjusted for socio-demographic variableslisted in table footnotes; all analyses were repeatedstratifying the sample by gender due to the significantinteraction effect observed in the relationship betweenoverweight or obesity and gender and combined beha-viors (p<0.05).ResultsThe proportion of the overall sample classified as over-weight or obese, engaging in S-MVPA and HST as separ-ate behaviors in the pooled sample was 59.4%, 46.2%and 43.1% respectively. Examination of combined activ-ity behaviors indicates that proportions of the overallsample classified as S-MVPA/LST, I-MVPA/LST, S-MVPA/HST, I-MVPA/HST was 26.1%, 29.8%, 20.0% and24.0% respectively (Table 1).Table 2 displays relationships between engagement invarious activity behaviors and the likelihood to be classi-fied as overweight or obese. Engagement in S-MVPAwas inversely associated with risk of overweight or obes-ity in the overall sample (OR=0.85, 95% CI. 0.78-0.93)and in females (OR=0.76, 95% CI. 0.67-0.86). Engage-ment in HST was positively associated with risk of over-weight or obesity in the overall sample (OR=1.38, 95%CI. 1.26-1.52), males (OR=1.41, 95% CI. 1.23-1.61) andin females (OR=1.40, 95% CI. 1.23-1.59). Examination ofcombined activity behaviors indicates that engagementin HST is positively associated with risk of overweight orobesity regardless of activity classification, and that themagnitude of association was greatest for those in the I-MVPA/HST category. This pattern of association waspresent in the overall sample, males and females(Table 2). Engagement in I-MVPA/LST was significantlyassociated with risk of overweight or obesity howeverthis was observed only in females.Table 3 displays that the proportion of the overallpopulation, males and females classified as overweightor obese significantly increased at the same rate over thestudy period. The proportion of the overall sample andfemales engaging in S-MVPA (OR=1.03, 95% CI. 1.01-1.05; OR=1.04, 95% CI. 1.01-1.07) also significantlyTable 2 Associations between physical activity, screen time, combined behavior categories and the likelihood to beclassified as overweight or obeseOverall Sample Males FemalesOR (95% CI) p OR (95% CI) p OR (95% CI) pPhysical ActivityInsufficient Physical Activity (I-MVPA) Referencea Referenceb ReferencecSufficient Physical Activity (S-MVPA) 0.85 (0.78-0.93) 0.001 0.99 (0.87-1.13) 0.852 0.76 (0.67-0.86) <0.001Screen ActivityLow Screen Time (LST) Referencea Referenceb ReferencecHigh Screen Time (HST) 1.38 (1.26-1.52) <0.001 1.41 (1.23-1.61) <0.001 1.40 (1.23-1.59) <0.001Combined BehaviorcS-MVPA/LST Referencea Referenceb ReferencecI-MVPA/LST 1.10 (0.98-1.24) 0.111 0.99 (0.84-1.17) 0.912 1.20 (1.01-1.42) 0.032S-MVPA/HST 1.29 (1.13-1.47) <0.001 1.37 (1.13-1.66) 0.001 1.23 (1.02-1.49) 0.028I-MVPA/HST 1.62 (1.42-1.84) <0.001 1.42 (1.18-1.71) <0.001 1.83 (1.53-2.19) <0.001a Adjusted for Gender, Age, Education, Employment Status, Year of Survey, Smoking Status. n = 7908.b Adjusted for Age, Education, Employment Status, Year of Survey, Smoking Status. n = 4079.c Adjusted for Age, Education, Employment Status, Year of Survey, Smoking Status. n = 3829.Duncan et al. BMC Public Health 2012, 12:1060 Page 4 of 9http://www.biomedcentral.com/1471-2458/12/1060increased. The proportion of the overall sample, malesand females engaging in HST increased significantly overthe study period (OR=1.03, 95% CI. 1.01-1.06; OR=1.04,95% CI.1.01-1.08; OR=1.03, 95% CI. 1.00-1.07). The pro-portion of the overall and female population classified asI-MVPA/LST declined over the study period (Table 3).Figure 1 displays the change in the proportion of theoverall population classified as overweight and obese, S-MVPA/LST, I-MVPA/LST, S-MVPA/HST and S-MVPA/LST by year of survey.Table 4 displays changes in the prevalence of over-weight or obesity within each behavior category duringthe study period. Tests of interaction effects betweeneach behaviour category and year of survey were not sta-tistically significant for any of the outcomes presented inTable 4 (p>0.05). In the overall sample the prevalence ofoverweight or obesity significantly increased in thosepeople classified as engaging in I-MVPA and S-MVPA(OR=1.05, 95% CI. 1.02-1.08; OR=1.04, 95% CI. 1.00-1.07), LST and HST (OR=1.03, 95% CI. 1.00-1.06;OR=1.05, 95% CI. 1.02-1.09) and I-MVPA/HST(OR=1.06, 95% CI. 1.01-1.11). In males the prevalence ofoverweight and obesity significantly increased in thoseclassified as engaging in S-MVPA (OR=1.06, 95% CI.1.01-1.10) and LST (OR=1.04, 95% CI. 1.00-1.09). Theprevalence of overweight or obesity significantlyincreased in those females classified as engaging in I-MVPA (OR=1.07, 95% CI. 1.02-1.11), HST (OR=1.07,95% CI. 1.02-1.12) and I-MVPA/HST (OR=1.10, 95% CI.1.03-1.17).DiscussionThis study examined relationships between physical ac-tivity, screen time, the combination of these behaviorsand the likelihood to be classified as overweight or obeseand also temporal trends in these outcomes during theperiod 2002–2008. In the overall sample and in females,engagement in S-MVPA was associated with a reducedlikelihood to be classified as overweight or obese, whileengagement in HST was associated with an increasedlikelihood to be classified as overweight or obese irre-spective of activity level measured by the Active AustraliaQuestionnaire. This pattern of associations is consistentwith previous research that examined similar behavioraland health outcomes [14,21]. In males, with the ex-ception of participation in S-MVPA, associations be-tween screen time and combined behaviors and thelikelihood to be classified as overweight or obese fol-lowed expected patterns [16,31]. The lack of associ-ation between S-MVPA and overweight or obesity inmales in the current study is both in agreement[35,36], and in contrast to previous studies [37,38], andmay be attributed to an positive energy imbalance,caused by energy intake exceeding energy expenditureTable 3 Trends in overweight and obesity, physical activity, screen time and combined behaviors during the period2002-2008Overall Sample Males FemalesOR (95% CI) p OR (95% CI) p OR (95% CI) pBody WeightHealthy Weight Referencea Referenced ReferencefOverweight or Obese 1.04 (1.02-1.06) 0.001 1.04 (1.01-1.07) 0.021 1.04 (1.01-1.07) 0.015Physical ActivityInsufficient Physical Activity (I-MVPA) Referenceb Referencee ReferencehSufficient Physical Activity (S-MVPA) 1.03 (1.01-1.05) 0.010 1.02 (0.99-1.05) 0.211 1.04 (1.01-1.07) 0.013Screen ActivityLow Screen Time (LST) Referenceb Referencee ReferencehHigh Screen Time (HST) 1.03 (1.01-1.06) 0.004 1.04 (1.01-1.08) 0.009 1.03 (1.00-1.07) 0.049Combined BehaviorS-MVPA/LST Referencec Referenceg ReferenceiI-MVPA/LST 0.97 (0.94-0.99) 0.019 0.98 (0.94-1.02) 0.327 0.95 (0.92-0.99) 0.019S-MVPA/HST 1.03 (1.00-1.06) 0.098 1.04 (1.00-1.09) 0.073 1.02 (0.98-1.07) 0.348I-MVPA/HST 1.00 (0.97-1.04) 0.782 1.02 (0.98-1.07) 0.334 0.99 (0.95-1.04) 0.763a Adjusted for Gender, Age, Education, Employment Status, Physical Activity, Smoking Status & Screen Time. n=7908.b Adjusted for Gender, Age, Education, Employment Status, & Smoking Status. n=7908.c Adjusted for Gender, Age, Education, Employment Status, & Smoking Status. n=7908.d Adjusted for Age, Education, Employment Status, Smoking Status, Physical Activity, & Screen Time. n=4079.e Adjusted for Age, Education, Employment Status, Smoking Status. n=4079.f Adjusted for Age, Education, Employment Status, Smoking Status, Physical Activity, & Screen Time. n=3829.g Reference Category is all other behavior categories. Adjusted for Age, Education, Employment Status, Smoking Status. n=4079.h Adjusted for Age, Education, Employment Status, Smoking Status. n=3829.i Reference Category is all other behavior categories. Adjusted for Age, Education, Employment Status, Smoking Status. n=3829.Duncan et al. BMC Public Health 2012, 12:1060 Page 5 of 9http://www.biomedcentral.com/1471-2458/12/1060even when males engage in S-MVPA [7,11]. The currentstudy does not include a measure of energy intake whichlimits the ability to examine this mechanism.While TV viewing is more broadly reflective of femalessedentary activity than males, we attempted to offset thisin the current study by incorporating a measure of com-puter use which is an activity that contributes more tomales overall sedentary time than females [32]. Despitethis, in females, associations between MVPA, screenbehaviors and overweight or obesity were observed in allcategories of combined behavior (I-MVPA/LST; S-MVPA/HST; I-MVPA/HST) whilst in males, associationsbetween combined behavior and overweight or obesitywere observed for selected behaviors (S-MVPA/HST;I-MVPA/HST). Several potential mechanisms may beattributed to this. The screen time and physical activitymeasures used in this study may better capture overallenergy expenditure in females compared to males. Alter-natively differences in dietary patterns may also contrib-ute to the differing associations observed. Thus futurestudies should consider using measures of physical activityand sedentary behavior that capture activity in all domainsand incorporate a measure of energy intake or dietary qual-ity. Withstanding these comments, the increased risk ofTable 4 Trends in overweight and obesity within behavior categories during the period 2002-2008Trend for Overweight & Obese within Behavior CategoriesOverall Sample Males Femalesn OR (95% CI) p n OR (95% CI) p n OR (95% CI) pPhysical ActivityInsufficient Physical Activity (I-MVPA) 4256 1.05 (1.02-1.08)a 0.002 2173 1.03 (0.99-1.08)b 0.154 2083 1.07 (1.02-1.11)b 0.007Sufficient Physical Activity (S-MVPA) 3652 1.04 (1.00-1.07)a 0.039 1906 1.05 (1.00-1.10)b 0.032 1746 1.02 (0.97-1.06)b 0.319Screen ActivityLow Screen Time (LST) 4426 1.03 (1.00-1.06)a 0.049 2260 1.04 (1.00-1.09)b 0.043 2166 1.01 (0.97-1.06)b 0.517High Screen Time (HST) 3482 1.05 (1.02-1.09)a 0.005 1819 1.03 (0.98-1.09)b 0.232 1663 1.07 (1.02-1.12)b 0.008Combined BehaviorS-MVPA/LST 2067 1.02 (0.98-1.07)a 0.264 1057 1.05 (0.98-1.11)b 0.156 1010 1.00 (0.94-1.06)b 0.957I-MVPA/LST 2359 1.04 (1.00-1.08)a 0.070 1203 1.05 (0.99-1.11)b 0.131 1156 1.03 (0.97-1.09)b 0.302S-MVPA/HST 1585 1.05 (0.99-1.10)a 0.089 849 1.06 (0.98-1.14)b 0.127 736 1.03 (0.96-1.11)b 0.354I-MVPA/HST 1897 1.06 (1.01-1.11)a 0.021 970 1.01 (0.94-1.08)b 0.891 927 1.10 (1.03-1.17)b 0.005a Adjusted for Gender, Age, Education, Employment Status & Smoking Status.b Adjusted for Age, Education, Employment Status & Smoking Status.0102030405060702002 2003 2004 2005 2006 2007 2008% of PopulationYear of SurveyOverweight or Obese S-MVPA/LSTI-MVPA/LST S-MVPA/HSTFigure 1 Trends in the prevalence of overweight and obesity and combined behaviors during the period 2002–2008.Duncan et al. BMC Public Health 2012, 12:1060 Page 6 of 9http://www.biomedcentral.com/1471-2458/12/1060overweight or obesity in individuals who engaged in HST(irrespective of measured MVPA) highlights the need forinterventions to target reductions in sedentary activitiesand increases in light, moderate and vigorous intensityphysical activity to maximize overall increases in energyexpenditure to reduce the risk of overweight or obesity.Increasing light intensity physical activity may be import-ant as it is likely that this is the behavior engaged inwhen sitting time is reduced [39] and increased light in-tensity physical activity is associated with improvedmetabolic health [40].Several studies have reported that the proportion ofthe population engaging in S-MVPA has remained stableor increased in Australian populations [4], including thispopulation [5], and populations from other countries[3,41]. As such the more unique aspect of this study,withstanding variations between individual years exam-ined (Table 1), is that it demonstrates that in the overallpopulation and in females, the proportion of the popula-tion classified as overweight or obese has increased overthe same time period and at a similar rate to theincreased proportion of the population engaging in suffi-cient MVPA and high screen time. Similar patterns ofchange were observed in males however the change inthe prevalence of S-MVPA over the study period wasnot statistically significant. Therefore the promisingchanges in the prevalence of sufficient MVPA should beviewed with cautious optimism. As it appears that somesegments of the population have responded to recentefforts to promote engagement in MVPA, and at thesame time also appear to spend increasing amounts oftime in sedentary activities. Still, these data providemuch needed information on the trends of these mul-tiple behaviors at a population level.The proportion of the population classified as over-weight or obese increased in most physical activity andscreen time groups when examined as separate behaviorgroups, and significantly increased only in the I-MVPA/HST group (overall sample and females) when combinedbehavior groups were examined. Furthermore, the rateof change in the combined activity group was margin-ally larger than observed when separate behaviorgroups were examined, although this was not signifi-cant in males. Although these conclusions must beinterpreted with caution as the data are not longitu-dinal, and there are inconsistent data surrounding theassociations between screen time and weight statuscompared to weight gain [15,42].Although interesting, findings in the current study aresubject to several limitations, including the use of self-report measures of MVPA, screen time and BMI. Self-reported BMI although practical in population basedstudies such as the current study has well acknowledgedlimitations [43] therefore future studies are encouragedto use objective measures of body composition to con-firm the pattern of results observed in the current study.Also the measures of screen time used may be not rep-resentative of the broader time spent in sedentary activ-ities [32]. Other limitations include the absence oflongitudinal data, a lack of data on actual sedentary be-havior and reliance on proxy measures of these beha-viors, the absence of a measure of sedentary activity intransport related activities and the absence of a measureof energy intake. Strengths of the study are the use ofconsistent methodology and survey instruments over thestudy period and the study sample size. The sample sizeof the study meant that even relatively small shifts in thepopulation prevalence of behaviors, overweight andobesity over the study period were statistically signifi-cant. However, the results do highlight importantchanges in behaviors at the population level.ConclusionsThe findings of this study support previous observationsthat high levels of time spent engaged in screen basedactivity is associated with overweight or obesity in crosssectional analyses, even when MVPA is considered[16,21]. It was also observed that the prevalence of over-weight or obesity appears to have increased at a similarmagnitude to the prevalence of the population that en-gage in sufficient MVPA and high levels of screen time.Finally, we found that in the overall sample and females,the prevalence of overweight or obesity increased overtime only in those who participated in insufficient levelsof MVPA and high levels of screen time. Greater under-standing of these relationships and trends over timerequires measures of actual sedentary behaviors, such assitting, that accurately capture the behaviors of bothmales and females and examination of these behaviors inboth population and cohort based samples.Competing interestsThe author declare that they have no competing interests.Authors’ contributionsMJD proposed the study concept, conducted statistical analyses and draftedthe original manuscript. CH assisted in statistical analyses. CC, CV and WKMprovided important intellectual feedback and critique of the study conceptand approach. All authors contributed to the drafting and editing of themanuscript and approved the final manuscript.AcknowledgementsThe authors would like to thank the staff of the Population ResearchLaboratory, CQUniversity and participants who have taken part in the CQSSduring the study period. Vandelanotte is supported by a National Health andMedical Research Council of Australia (#519778) and National HeartFoundation of Australia (#PH 07B 3303) post-doctoral research fellowship.The Central Queensland Social Survey is an annual omnibus survey fundedby the Institute for Health and Social Science Research (IHSSR) and isconducted by the Population Research Laboratory at CQUniversity Australia.Author details1CQUniversity, Institute for Health and Social Science Research, Centre forPhysical Activity Studies, Rockhampton, Bld 18, CQUniversity Australia,Duncan et al. BMC Public Health 2012, 12:1060 Page 7 of 9http://www.biomedcentral.com/1471-2458/12/1060Rockhampton, QLD 4702, Australia. 2Faculty of Health and SocialDevelopment, University of British Columbia, Kelowna, Canada. 3Faculty ofPhysical Education and Recreation, University of Alberta, Edmonton, Canada.Received: 26 January 2012 Accepted: 30 November 2012Published: 8 December 2012References1. Bauman AE: Updating the evidence that physical activity is good forhealth: an epidemiological review 2000–2003. J Sci Med Sport 2004,7(1 Suppl):6–19.2. Brown W, Bauman A, Chey T, Trost S, Mummery K: Comparison of surveysused to measure physical activity. Aust N Z J Public Health 2004,28(2):128–134.3. Brownson RC, Boehmer TK, Luke DA: Declining rates of physical activity inthe United States: what are the contributors? Annu Rev Public Health 2005,26:421–443.4. Chau J, Smith BJ, Bauman A, Merom D, Eyeson-Annan M, Chey T,Farrell L: Recent trends in physical activity in New South Wales. Isthe tide of inactivity turning? Aust N Z J Public Health 2008, 32(1):82–85.5. Vandelanotte C, Duncan MJ, Caperchione C, Hanley C, Mummery WK:Physical activity trends in Queensland (2002 to 2008): are womenbecoming more active than men? Aust N Z J Public Health 2010,34(3):248–254.6. Knuth AG, Hallal PC: Temporal trends in physical activity: a systematicreview. J Phys Act Health 2009, 6(5):548–559.7. Bauman A, Allman-Farinelli M, Huxley R, James WP: Leisure-time physicalactivity alone may not be a sufficient public health approach to preventobesity–a focus on China. Obes Rev 2008, 9(Suppl 1):119–126.8. Australian Bureau of Statistics: Australian Social Trends 2007. AustralianBureau of Statistics: Overweight and obesity. In. Canberra; 2007.9. Flegal KM, Carroll MD, Ogden CL, Curtin LR: Prevalence and trends inobesity among US adults, 1999–2008. JAMA, 303(3):235–241.10. Stamatakis E, Ekelund U, Wareham NJ: Temporal trends in physical activityin England: the Health Survey for England 1991 to 2004. Prev Med 2007,45(6):416–423.11. Hill JO, Wyatt HR, Reed GW, Peters JC: Obesity and the environment:where do we go from here? Science 2003, 299(5608):853–855.12. Pate RR, O'Neill JR, Lobelo F: The evolving definition of "sedentary".Exerc Sport Sci Rev 2008, 36(4):173–178.13. Clark BK, Sugiyama T, Healy GN, Salmon J, Dunstan DW, Owen N: Validityand reliability of measures of television viewing time and othernon-occupational sedentary behaviour of adults: a review. Obes Rev 2009,10(1):7–16.14. Hu FB, Leitzmann MF, Stampfer MJ, Colditz GA, Willett WC, Rimm EB:Physical activity and television watching in relation to risk for type 2diabetes mellitus in men. Arch Intern Med 2001, 161(12):1542–1548.15. Hu FB, Li TY, Colditz GA, Willett WC, Manson JE: Television watching andother sedentary behaviors in relation to risk of obesity and type 2diabetes mellitus in women. JAMA 2003, 289(14):1785–1791.16. Vandelanotte C, Sugiyama T, Gardiner P, Owen N: Associations ofleisure-time internet and computer use with overweight and obesity,physical activity and sedentary behaviors: cross-sectional study. J MedInternet Res 2009, 11(3):e28.17. Owen N, Bauman A, Brown W: Too Much Sitting. Br J Sports Med: A Noveland Important Predictor of Chronic Disease Risk? 2008.18. Hamilton MT, Healy GN, Dunstan DW, Zderic TW, Owen N: Too littleexercise and too much sitting: inactivity physiology and the need fornew recommendations on sedentary behavior. Current Cardiovascular RiskReports 2008, 2(4):292–298.19. Church TS, Thomas DM, Tudor-Locke C, Katzmarzyk PT, Earnest CP, RodarteRQ, Martin CK, Blair SN, Bouchard C: Trends over 5 decades in U.S.Occupation-related physical activity and their associations with obesity.PLoS One 2011, 6(5):19657.20. Chau JY, Merom D, Grunseit A, Rissel C, Bauman AE, van der PloegHP: Temporal trends in non-occupational sedentary behavioursfrom Australian Time Use Surveys 1992, 1997 and 2006.International Journal of Behavioral Nutrition and Physical Activity 2012,9(1):76.21. Sugiyama T, Healy GN, Dunstan DW, Salmon J, Owen N: Joint associationsof multiple leisure-time sedentary behaviours and physical activity withobesity in Australian adults. Int J Behav Nutr Phys Act 2008, 5:35.22. Ainsworth BE, Wilcox S, Thompson WW, Richter DL, Henderson KA:Personal, social, and physical environmental correlates of physicalactivity in African-American women in South Carolina. Am J Prev Med2003, 25(3 Suppl 1):23–29.23. Badland HM, Schofield GM: Health Associations with Transport-RelatedPhysical Activity and Motorized Travel to Destinations. Int J SustainableTransportation 2008, 2(2):77–90.24. Australian Institute of Health and Welfare: The active Australia survey: A guideand manual for implementation, analysis and reporting. Canberra:AIHW; 2003.25. Brown WJ, Bauman A, Trost S, Mummery WK, Owen N: Test-retestreliability of four physical activity measures used in population surveys.Journal of Science and Medicine in Sport 2004, 7(2):205–215.26. Timperio A, Salmon J, Bull F, Rosenburg M: Validation of adult physicalactivity questions for use in Australia population surveys. Canberra:Department of Health and Ageing; 2002. unpublished report.27. Department of Health and Aged Care: National physical activity guidelines forAustralians. Canberra: DHAC; 1999.28. Dunstan DW, Salmon J, Owen N, Armstrong T, Zimmet PZ, Welborn TA,Cameron AJ, Dwyer T, Jolley D, Shaw JE: Physical activity and televisionviewing in relation to risk of undiagnosed abnormal glucose metabolismin adults. Diabetes Care 2004, 27(11):2603–2609.29. Dunstan DW, Salmon J, Healy GN, Shaw JE, Jolley D, Zimmet PZ, Owen N:Association of television viewing with fasting and 2-h postchallengeplasma glucose levels in adults without diagnosed diabetes. DiabetesCare 2007, 30(3):516–522.30. Bertrais S, Beyeme-Ondoua JP, Czernichow S, Galan P, Hercberg S,Oppert JM: Sedentary behaviors, physical activity, and metabolicsyndrome in middle-aged French subjects. Obes Res 2005, 13(5):936–944.31. Dunstan DW, Salmon J, Owen N, Armstrong T, Zimmet PZ, Welborn TA,Cameron AJ, Dwyer T, Jolley D, Shaw JE: Associations of TV viewing andphysical activity with the metabolic syndrome in Australian adults.Diabetologia 2005, 48(11):2254–2261.32. Sugiyama T, Healy GN, Dunstan DW, Salmon J, Owen N: Is televisionviewing time a marker of a broader pattern of sedentary behavior?Ann Behav Med 2008, 35(2):245–250.33. Davies CA, Vandelanotte C, Duncan MJ, van Uffelen JG: Associations ofphysical activity and screen-time on health related quality of life inadults. Prev Med 2012, 55(1):46–49.34. Banks E, Jorm L, Rogers K, Clements M, Bauman A: Screen-time, obesity,ageing and disability: findings from 91 266 participants in the 45 andUp Study. Public Health Nutr 2011, 14(1):34–43.35. Rainwater DL, Mitchell BD, Comuzzie AG, VandeBerg JL, Stern MP, MacCluerJW: Association among 5-year changes in weight, physical activity, andcardiovascular disease risk factors in Mexican Americans. Am J Epidemiol2000, 152(10):974–982.36. Seo DC, Li K: Leisure-time physical activity dose–response effects onobesity among US adults: results from the 1999–2006 National Healthand Nutrition Examination Survey. J Epidemiol Community Health 2010,64(5):426–431.37. Salmon J, Bauman A, Crawford D, Timperio A, Owen N: The associationbetween television viewing and overweight among Australian adultsparticipating in varying levels of leisure-time physical activity. Int J ObesRelat Metab Disord 2000, 24(5):600–606.38. Lahti-Koski M, Pietinen P, Heliovaara M, Vartiainen E: Associations of bodymass index and obesity with physical activity, food choices, alcoholintake, and smoking in the 1982–1997 FINRISK Studies. Am J Clin Nutr2002, 75(5):809–817.39. Owen N, Healy GN, Matthews CE, Dunstan DW: Too much sitting: thepopulation-health science of sedentary behavior. Exercise and sportsciences reviews 2010, 38(3):105.40. Healy GN, Dunstan DW, Salmon J, Cerin E, Shaw JE, Zimmet PZ, OwenN: Objectively measured light-intensity physical activity isindependently associated with 2-h plasma glucose. Diabetes Care2007, 30(6):1384–1389.41. Juneau CE, Potvin L: Trends in leisure-, transport-, and work-relatedphysical activity in Canada 1994–2005. Prev Med 2010, 51(5):384–386.Duncan et al. BMC Public Health 2012, 12:1060 Page 8 of 9http://www.biomedcentral.com/1471-2458/12/106042. Van Uffelen JG, Watson MJ, Dobson AJ, Brown WJ: Sitting time isassociated with weight, but not with weight gain in mid-aged Australianwomen. Obesity (Silver Spring) 2010, 18(9):1788–1794.43. Strommel M, Schoenborn CA: Accuracy and usefulness of BMI measuresbased on selfreported weight and height: findings from the NHANES &NHIS 2001–2006. BMC Public Health 2009, 9:421.doi:10.1186/1471-2458-12-1060Cite this article as: Duncan et al.: Temporal trends in and relationshipsbetween screen time, physical activity, overweight and obesity. BMCPublic Health 2012 12:1060.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistributionSubmit your manuscript at www.biomedcentral.com/submitDuncan et al. BMC Public Health 2012, 12:1060 Page 9 of 9http://www.biomedcentral.com/1471-2458/12/1060


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items