Preventive Medicine 54 (2012) 61–64

Contents lists available at SciVerse ScienceDirect

Preventive Medicine

j ourna l homepage: www.e lsev ie r .com/ locate /ypmed

Brief Original Report

Socio-demographic characteristics of children experiencing socioeconomicdisadvantage who meet physical activity and screen-time recommendations: TheREADI study

Clare Hume ⁎, Jo Salmon, Jenny Veitch, Eoin O'Connell, David Crawford, Kylie BallCentre for Physical Activity and Nutrition Research, Deakin University, Burwood, Victoria, Australia

⁎ Corresponding author at: Centre for Physical ActivityUniversity, 221 Burwood Hwy, Burwood 3125, Victoria, A

E-mail address: clare.hume@deakin.edu.au (C. Hume

0091-7435/$ – see front matter © 2011 Published by Eldoi:10.1016/j.ypmed.2011.10.019

a b s t r a c t

a r t i c l e i n f o

Available online 3 November 2011

Keywords:Socioeconomic disadvantageUrban/ruralAccelerometerWeight status

Objective: To identify socio-demographic characteristics of children from socioeconomically disadvan-taged neighbourhoods who meet physical activity and screen recommendations.

Method: Children aged 5–12 years (n=373; 45% boys) were recruited in 2007 from socioeconomically disad-vantaged urban and rural areas of Victoria, Australia. Children's physical activity, height andweightwere objective-lymeasured.Mothers reported their highest level of education, and proxy-reported their child's usual screen-time.Odds ratios (OR) and 95% confidence intervals (95% CI) examinedodds ofmeeting physical activity (>60 minutes/

day) and screen (≤120 minutes/day) recommendations according to socio-demographic characteristics.

Results: Approximately 84% of children met physical activity and 43% met screen recommendations. Age wasinversely associated with odds of meeting physical activity and screen recommendations, and overweight/obesestatus was associated with lower odds of meeting screen recommendations (boys: OR=0.39, 95%CI=0.16–0.95;girls: OR=0.47, 95%CI=0.26–0.83). Among boys, living in a rural area was positively associated with meetingscreen recommendations (OR=3.08, 95%CI=1.42–6.64). Among girls, high levels of maternal education werepositively associated with meeting screen recommendations (OR=2.76, 95%CI=1.33–5.75).

Conclusion: Specific socio-demographic characteristics were associated with odds of meeting physical activityand screen recommendations. Identifying factors associated with such ‘resilience’ among this group may provideimportant learnings to inform future physical activity promotion initiatives.

© 2011 Published by Elsevier Inc.

Introduction

The health benefits of children's physical activity are well known(Biddle et al., 2004), and evidence of adverse health outcomes associ-atedwith children's participation in sedentary behaviours is emerging(Rey-Lopez et al., 2008). Health authorities recommend that childrenspend 60 minutes ormore inmoderate- to vigorous-intensity physicalactivity (MVPA) everyday (Commonwealth Department of Health andAgeing, 2004), and spend no more than 2 hours/day in screen-basedpursuits for entertainment (Commonwealth Department of Healthand Ageing, 2004). Importantly, much of the evidence suggests thatcompliance with these recommendations is poor (CommonwealthDepartment of Health and Ageing, 2004; Nader et al., 2008; Riddochet al., 2007; Stamatakis et al., 2009).

There is some evidence of associations between socioeconomic disad-vantage, physical activity and screen-time among children, regardless ofwhether individual-level (e.g. household income or maternal education)(Ball et al., 2009; Nader et al., 2008; Singh et al., 2009) or area-level

and Nutrition Research, Deakinustralia.).

sevier Inc.

measures (e.g. area deprivation) are used (Brodersen et al., 2007; Singhet al., 2009). It is therefore important to examine these behavioursamong children experiencing socioeconomic disadvantage. It is also im-portant to recognise that some children in that group are physically ac-tive, and some do participate in very little screen-time (Booth et al.,2006). These children could be described as ‘resilient’; that is, they appearto defy the increased odds of inactivity and sedentary behaviours associ-atedwith socioeconomic disadvantage (Ball and Crawford, 2006). Identi-fying factors associated with such ‘resilience’ may provide importantinformation about where future research should be targeted. The aim ofthis study was to identify the socio-demographic characteristics of chil-dren who meet physical activity and screen recommendations despiteliving in socioeconomically disadvantaged neighbourhoods.

Methods

Procedure

This study drew on data from the baseline assessment of the READI (Re-silience for Eating and Activity Despite Inequality) study, conducted betweenJuly 2007 and June 2008. The methods employed in that study have beenpublished previously (Cleland et al., 2010). Briefly, 40 urban and 40 ruralsuburbs in the lowest tertile of socioeconomic disadvantage (the most

Table 1Socio-demographic and behavioral characteristics of children in the READI studya.

Boys (n=167) Girls (n=206)

Age (years; mean, SD) 9.2 (2.1) 9.6 (2.0)Geographic location (%)Urban 30 30Rural 70 70Child's zBMI (Body Mass Index)[mean (SD)]b

0.6 (0.9) 0.5 (0.9)

Child's weight status (%)Not overweight 74 69Overweight 17 20Obese 9 11Maternal education (%)Low 25 25Medium 50 46High 25 29Moderate to vigorous physicalactivity [min/day] [mean (SD)]

195.0 (68.7)*** 156.4 (62.9)

Screen-time [min/day] [mean (SD)] 160.0 (97.1) 141.9 (83.7)Meeting physical activityrecommendations (%)

89.2* 79.6

Meeting screen-timerecommendations (%)

37.7 47.1

*pb0.05, ***pb0.001 between girls and boys.a Recruited in 2007/2008 from urban and rural areas of Victoria, Australia.b Standardised z-score of BMI.

62 C. Hume et al. / Preventive Medicine 54 (2012) 61–64

disadvantaged areas) (Australian Bureau of Statistics, 1998) were randomlyselected. Women aged between 18 and 45 years living in each suburb(n=150) were selected from the electoral roll to receive a survey(n=11,940). Consent was received from 4934 women (41% response rate).Limited information is available regarding non-respondents; however ahigher response rate was seen among rural compared to urban women(39% vs 34%). Women with a child aged 5–12 years were approached to par-ticipate, with 771 consenting (53% response rate). Mothers were mailed asurvey and a reply paid envelope, the child's school or home was visited formeasurement.

Measures

.Socio-demographic variablesMothers reported their highest level of education (low=no formal qual-

ifications, year 10 or equivalent; medium=Year 12 or equivalent, trade/ap-prenticeship, certificate/diploma; high=university degree or higher) andtheir child's date of birth. Urban or rural residence was determined duringsampling.

.Child's height and weightDuring the home/school visit, children's height (to the nearest 0.1 cm)

and weight (to the nearest 0.1 kg) were measured using portable stadi-ometers and scales. The average of two consecutive measures was calculated.From this, children's age- and sex-adjusted body mass index z-scores werederived (zBMI) and children's weight status (not overweight, overweight/obese) was calculated (Cole et al., 2000).

.Screen-timeMothers proxy-reported their child's TV viewing, computer and electron-

ic game use in a typical week and total average minutes/day was calculated(Salmon et al., 2006). Children who met current Australian screen-time rec-ommendations of ≤120 minutes/day were determined.

.Moderate-to-vigorous intensity physical activityDuring the home/school visit, children were fitted with a Manufactur-

ing Technologies Inc. (MTI) accelerometer (Actigraph Model AM7164-2.2C), worn on their right hip for eight consecutive days during wakinghours. Inclusion criteria were at least four valid days (10–18 hours ofwear time) including at least one weekend day. Average minutes/day ofMVPA was calculated (Trost et al., 1998) then adjusted for proportion ofwear time. Children who met the current Australian physical activity rec-ommendations (≥60 minutes/day MVPA) were determined.

Statistical analyses

Stata/SE version 10.1 was used to perform logistic regression analyses toexamine the odds of meeting recommendations for physical activity andscreen-time, according to socio-demographic characteristics. Factors associ-ated with the outcomes in bivariable analyses were entered into multivari-able regression models. Analyses were stratified by sex, and all modelsadjusted for clustering by suburb (the sampling unit).

Results

After exclusions and missing data, the final sample comprised 373(167 boys and 206 girls) children with complete data. The samplesocio-demographic characteristics and time spent in MVPA andscreen-time are presented in Table 1.

Table 2 shows the results of multivariable logistic regression ana-lyses examining the socio-demographic factors associated with oddsof meeting physical activity and screen recommendations. Amongboys and girls, age was inversely associated with meeting physical ac-tivity and screen recommendations. Compared to urban boys, ruralboys had higher odds of meeting screen recommendations. Girlswhose mothers had high levels of education had higher odds of meet-ing screen recommendations than other girls. Overweight boys andgirls were less likely to meet screen recommendations than non-overweight children.

Discussion

This study examined the socio-demographic characteristics ofchildren experiencing socioeconomic disadvantage who met physicalactivity and screen recommendations. Approximately 84% of childrenmet physical activity recommendations but only 43% met screen rec-ommendations. This study identified several important factors; spe-cifically age, rural residence, maternal education and weight status.These findings identify key target groups for whom further study ofmodifiable characteristics may be valuable, particularly in relationto sedentary behaviours.

Consistentwith the present findings, age-related cross-sectional dif-ferences in physical activity have been reported previously (Sallis,2000), and this finding was extended to screen-time. This is interestingas existing evidence of that association ismixed (Hoyos Cillero and Jago,2010). It is plausible that as children get older, time spent using thecomputer for homework increases, thus accounting for the increasedtotal screen-time. Screen-time also showed associations among ruralboys, and although data examining sedentary behaviours among ruralchildren are scarce; one previous study showed similar findings(Booth et al., 2006). Poorer access to screen-based technologies (e.g.the internet) in rural areas (Australian Bureau of Statistics, 2008) maycontribute to these differences. Given the limited research that has ex-amined screen-based behaviour in this population, further work iswarranted.

Findings from the current study in relation toweight status (Marshallet al., 2004) and maternal education (Ball et al., 2009; Hoyos Cillero andJago, 2010) are also consistentwith previous research. These associationsare particularly significant when considered in the context of this study,as all participants were recruited from socioeconomically disadvantagedareas.

The cross-sectional design limits inferences about causality, andthe measure of screen-time was proxy-reported; although the psy-chometric properties were adequate (Salmon et al., 2006). Addi-tionally, the measure of screen-time may not completely captureall components of sedentary behaviour. However a key aim of thisstudy was to examine whether children met screen-time recom-mendations, and the behaviours measured are common screen-

Table 2Adjusted odds ratios (OR) and 95% confidence intervalsa (95%CI) for meeting recommendations according to socio-demographic characteristics for boys and girlsb.

Meeting physical activityrecommendations OR(95%CI) [n: yes/no]c

Meeting screen-timerecommendations OR(95%CI) [n: yes/no]c

BOYS (n=167)Age 0.58 (0.42, 0.82)** 0.80 (0.67, 0.96)*

Location of residenceUrban (ref) 1 [n=42/8] 1 [n=11/39]Rural 2.07 (0.80, 5.36) [n=107/10) 3.08 (1.42, 6.64)**

[n=52/65]Child's weight statusNot overweight (ref) 1 [n=114/9] 1 [n=54/69]Overweight/Obese 0.40 (0.13, 1.26) [n=35/9] 0.39 (0.16, 0.95)*

[n=9/35]Maternal educationLow (ref) 1 [n=36/6] 1 [n=12/30]Medium 0.89 (0.31, 2.55) [n=73/11] 1.45 (0.51, 4.08)

[n=34/50]High 4.73 (0.35, 64.04) [n=40/1] 1.38 (0.47, 4.06)

[n=17/24]GIRLS (n=206)Age 0.39 (0.30, 0.51)*** 0.73 (0.63, 0.84)***

Location of residenceUrban (ref) 1 [n=47/15] 1 [n=28/34]Rural 1.62 (0.66, 3.97) [n=117/27] 1.06 (0.52, 2.16)

[n=69/75]Child's weight statusNot overweight (ref) 1 [n=111/30] 1 [n=75/66]Overweight/Obese 1.05 (0.43, 2.55) [n=53/12] 0.47 (0.26, 0.83**)

[n=22/43]Maternal educationLow (ref) 1 [n=40/11] 1 [n=19/32]Medium 0.60 (0.19, 1.88) [n=71/23] 0.96 (0.48, 1.92)

[n=37/57]High 0.84 (0.25, 2.88) [n=53/8] 2.76 (1.33, 5.75)**

[n=41/20]

*pb0.05 **pb0.01 ***pb0.001.a Results of multivariable logistic regression analyses adjusting for other variables significantly associated with the outcomes in bivariable analyses.b Recruited in 2007/2008 from urban and rural areas of Victoria, Australia.c The number of participants to compute the crude ORs.

63C. Hume et al. / Preventive Medicine 54 (2012) 61–64

based entertainment pastimes. The fact that mothers responded toa mail-out survey combined with the modest response rate re-ceived may have introduced some selection bias; however we feelthis is less problematic in this study examining associations, thanwould be the case in a study examining population prevalences. Afurther limitation is the wide age range of children in the study(5–12 years) as factors associated with physical activity andscreen-time may differ during these years. Strengths include theuse of objective measures of children's physical activity and weightstatus, and the inclusion of children living in urban and rural areas.

Conclusion

This study identified characteristics of resilient children that may beimportant to examine in future researchwhen trying to understand be-haviour andwhen designing interventions for key target groups. Futureresearch should explore the social and physical environments in whichthese children live to provide a better understanding of the overall char-acteristics of resilient children.

Conflict of interest statement

The authors declare no conflict of interest.

Acknowledgements

This study was funded by a National Health and Medical ResearchCouncil (NHMRC) Strategic Award. We gratefully acknowledge thecontributions of the Project Manager Dr Michelle Jackson and thestudy participants. Clare Hume and Jenny Veitch were supported by

National Heart Foundation of Australia (NHFA) post-doctoral re-search fellowships, Jo Salmon by an NHFA and Sanofi-Aventis careerdevelopment award, David Crawford by public health research fel-lowships from the Victorian Health Promotion Foundation and KylieBall by an NHMRC senior research fellowship.

References

Australian Bureau of Statistics, 1998. 1996 Census of Population and Housing: Socio-Economic Indexes for Areas, Catalogue No. 2039. Australian Bureau of Statistics,Canberra.

Australian Bureau of Statistics, 2008. Patterns of internet access in Australia, 2006. Cat.No. 8146.0.55.001. Australian Bureau of Statistics, Canberra.

Ball, K., Crawford, D., 2006. Socio-economic factors in obesity: a case of slim chance in afat world? Asia Pac. J. Clin. Nutr. 15 (Suppl.), 15–20.

Ball, K., et al., 2009. Socioeconomic position and children's physical activity and sedentarybehaviors: longitudinal findings from the CLAN study. J. Phys. Activ. Health 6, 289–298.

Biddle, S., et al., 2004. Health-enhancing physical activity and sedentary behaviour inchildren and adolescents. J. Sports Sci. 22, 679–701.

Booth, M., et al., 2006. NSW Schools Physical Activity and Nutrition Survey (SPANS)2004: Full report. NSW Department of Health, Sydney.

Brodersen, N.H., et al., 2007. Trends in physical activity and sedentary behaviour in ad-olescence: ethnic and socioeconomic differences. Br. J. Sports Med. 41, 140–144.

Cleland, V., et al., 2010.Urban–rural comparison ofweight status amongwomenand childrenliving in socioeconomically disadvantaged neighbourhoods. Med. J. Aust. 192, 137–140.

Cole, T.J., et al., 2000. Establishing a standard definition for child overweight and obesityworldwide: international survey. BMJ 320, 1240–1243.

Commonwealth Department of Health and Ageing, 2004. Australia's Physical ActivityRecommendations for 5–12 year olds. Commonwealth of Australia, Canberra.

Hoyos Cillero, I., Jago, R., 2010. Systematic review of correlates of screen-viewingamong young children. Prev. Med. 51, 3–10.

Marshall, S.J., et al., 2004. Relationships betweenmedia use, body fatness andphysical activityin children and youth: a meta-analysis. Int. J. Obes. Relat. Metab. Disord. 28, 1238–1246.

Nader, P.R., et al., 2008. Moderate-to-vigorous physical activity from ages 9 to 15 years.J. Am. Med. Assoc. 300, 295–305.

64 C. Hume et al. / Preventive Medicine 54 (2012) 61–64

Rey-Lopez, J.P., et al., 2008. Sedentary behaviour and obesity development in childrenand adolescents. Nutr. Metab. Cardiovasc. Dis. 18, 242–251.

Riddoch, C.J., et al., 2007. Objective measurement of levels and patterns of physical activity.Arch. Dis. Child. 92, 963–969.

Sallis, J.F., 2000. Age-related decline in physical activity: a synthesis of human and animalstudies. Med. Sci. Sports Exerc. 32, 1598–1600.

Salmon, J., et al., 2006. Television viewing habits associated with obesity risk factors: asurvey of Melbourne schoolchildren. Med. J. Aust. 184, 64–67.

Singh, G.K., et al., 2009. Prevalence and correlates of state and regional disparities invigorous physical activity levels among US children and adolescents. J. Phys.Activ. Health 6, 73–87.

Stamatakis, E., et al., 2009. Television viewing and other screen-based entertainment inrelation to multiple socioeconomic status indicators and area deprivation: theScottish Health Survey 2003. J. Epidemiol. Community Health 63, 734–740.

Trost, S.G., et al., 1998. Validity of the computer science and applications (CSA) activitymonitor in children. Med. Sci. Sports Exerc. 30, 629–633.