In the past few decades, rates of overweight and obesity and related metabolic and cardiovascular risk factors in children have steadily increased worldwide [1–4]. Physical activity plays an important role in the prevention of metabolic and cardiovascular health risk factors in children . Increasingly, evidence suggests that sedentary behaviors, such as prolonged periods of television viewing, electronic games, and computer use (collectively called screen-time) may adversely affect children's weight status, independent of physical activity participation [6, 7]. While these screen behaviors are most commonly performed during children's leisure-time there are many opportunities throughout the day for children to be sedentary (e.g., being driven to school, sitting in class, and sedentary homework). Further, while some longitudinal evidence suggests that increases in the time children spend in sedentary behaviors seem to be off-set by corresponding decreases in physical activity during the primary school years , other studies have reported independent changes in these behaviors over time suggesting that they should be viewed as separate rather than converse constructs .
There is emerging evidence that not just screen time, but total sedentary time may be detrimental to children's health. A cross-sectional study of 208 Portuguese children (mean age 9.8 years) found positive associations between accelerometer-measured sedentary time (defined as < 500 Actigraph counts per minute [cpm]) and insulin resistance, and inverse associations between moderate- to vigorous-intensity physical activity (MVPA; ≥2, 000 cpm) and insulin resistance . A study of 111 US children (aged 3-8 years), however, found no cross-sectional associations between time spent sedentary (< 100 cpm) and systolic or diastolic blood pressure (BP) . Nevertheless, that study reported that children in the highest tertile for proxy-reported television viewing time (approximately 155 mins/day) were significantly more likely to have higher systolic and diastolic BP compared with children in the lowest tertile (approximately 8 mins/day).
Observational evidence from studies among adults suggests that the manner in which time spent sedentary is accumulated may also be detrimental to health. For example, a cross-sectional study with 168 Australian adults (mean age 53 years) found that independent of MVPA levels, those with less frequent interruptions to accelerometer-measured sedentary time (≥100 cpm) with light-intensity physical activity had less favourable health profiles (waist circumference, body mass index, triglycerides, 2-hour plasma glucose) compared to those with more frequent interruptions . Interestingly, the average duration of light-intensity breaks was less than five minutes, suggesting that even brief interruptions to time spent sedentary may be beneficial to health. To our knowledge, no observational or experimental studies have examined the association of interruptions to sedentary or sitting time and health in children, nor has the role of light-intensity physical activity and children's health been previously studied.
While few intervention studies have examined the effectiveness of strategies to reduce children's overall sedentary time, several review papers have summarized the effectiveness of interventions to reduce children's screen time [6, 13, 14]. While this evidence suggests these strategies (delivered primarily through school-based curriculum), have positive effects on children's weight and have successfully reduced TV viewing, as noted earlier, there are many opportunities to be active throughout the day both at school and at home  and few if any of these interventions to reduce children's screen time have resulted in corresponding increases in physical activity.
Several studies have reported significant positive effects on children's physical activity in the school setting by targeting the school curriculum or through changes in the school environment [16–19]; however, few studies report on intervention effects on children's sedentary time. A recent experimental study by Benden et al. among children in four classes in Central Texas introduced standing desks into classrooms and found that after 12 weeks all children were standing for 75% of the time . However, the intervention only targeted energy expenditure at school and did not incorporate strategies to increase energy expenditure or reduce sedentary behavior outside of school hours. A further challenge with this type of intervention is whether the aim is to reduce children's sedentary time, increase physical activity or both. In a meta-analysis of intervention studies that aimed to promote young people's physical activity or reduce screen time, pooled effect sizes of 0.12 and -0.29 respectively were reported . The authors concluded that strategies to reduce sedentary behavior appeared to be more effective than strategies to increase physical activity. However, the efficacy of strategies to increase physical activity and reduce sedentary behavior separately and in combination has not been examined.
Ecological models suggest that settings-based approaches may be an effective method for intervening with children's health behaviors [20, 21]. Interventions that target places and contexts in which large numbers of children are sedentary or active are likely to have a greater public health impact than approaches that involve one-on-one program delivery. In addition, an important aspect in the development of effective and efficacious behavioral interventions is the use of a theoretical framework [22, 23]. The use of behavioral theory helps guide the development of strategies that are most likely to result in changes in behavior through targeting the key mechanisms or mediating constructs of change [24, 25]. Commonly employed theories in children's physical activity and sedentary behavior intervention studies include: social cognitive theory ; theory of planned behavior ; and behavioral choice theory . A limitation of many of these theories is the focus on intrapersonal factors, which on the one hand are important for targeting change at the individual level, but are less useful when targeting changes at the population level. More recently, but less frequently, ecological models such as the social ecological model of health promotion  and the family-based ecological systems theory  have also been employed in interventions to promote children's physical activity with mixed success [30, 31].
Very few studies, even those that report use of behavioral theory in the design of their intervention, examine the mediators or mechanisms of behavior change. Several reviews of mediators of physical activity interventions in children and youth have identified key mediators to target including: self-efficacy; behavioral capability; perceived social support; physical activity knowledge and beliefs; and enjoyment of or preference for physical activity [32–34]. Just two studies have examined possible mediators of change in sedentary behaviors such as television viewing and computer use in young people [35, 36]. The DOiT study was an obesity prevention intervention based on the theory of planned behavior and habit strength theory  that aimed to improve dietary and physical activity habits as well as reduce sedentary time of Dutch adolescents . In that study there were no mediating effects of attitude, subjective norms (i.e. the degree to which an individual is inclined to agree with the expectation of other important persons' opinions, normative beliefs), behavioral control or habit strength on youth screen time. Based on the self-determination theory  and the theory of meanings of behavior , the Get Moving! program was a media-based intervention delivered via the school setting that aimed to increase physical activity and decrease sedentary behaviors in predominantly Latino middle school girls in California, USA . The authors found a non-significant trend for a mediating effect of intrinsic motivation to be physically active on television viewing time. No other mediating effects were observed.
It is therefore important that intervention studies not only target key mediators that lie on the behavior change pathway, but that these pathways are then tested statistically. This will ensure a better understanding of why an intervention worked or not and will further inform the utility of behavior change theories. Another often-overlooked aspect of children's health behavior change interventions is the economic cost of program delivery. Not only is it important to test whether an intervention works and why, it is also critical that it is cost effective. Cost-effectiveness analysis combines effectiveness and cost data to show whether an intervention represents 'value for money', with results expressed as incremental cost-effectiveness ratios. A range of standard methods are available to guide economic evaluation of an intervention program [40, 41]. For example, the Assessing Cost Effectiveness (ACE)-Obesity study examined the economic evaluation of thirteen interventions which targeted unhealthy weight gain in children and adolescents [42, 43]. While the cost-effectiveness of interventions varied greatly, the most cost-effective strategies included 'Reduction of TV advertising of high fat and/or high sugar foods and drinks to children', 'Laparoscopic adjustable gastric banding' and the 'multi-faceted school-based programme with an active physical education component'. Further research is required to identify the cost-effectiveness of strategies to reduce children's sedentary behavior and promote physical activity in the school and home settings.
This proposal builds on our program of research [42, 44, 45] aimed at identifying effective and cost-effective strategies that positively influence children's health behavior and translate to improved health outcomes. This paper presents a summary of the Transform-Us! intervention including its aims, development, intervention methods and assessment protocols.
The primary aim of the Transform-Us! study is to determine whether an 18-month, behavioral and environmental intervention in the school and family settings results in higher levels of physical activity and lower rates of sedentary behavior among 8-9 year old children compared with usual practice (post-intervention and 12-months follow-up). The secondary aims are to determine the independent and combined effects of PA and SB on children's cardio-metabolic health risk factors; identify the factors that mediate the success of the intervention; and determine whether the intervention is cost-effective.
Study Protocol Overview
Transform-Us! is a four-arm cluster randomized controlled trial with primary schools in Melbourne, Australia being the unit of randomization. The intervention will run for approximately 18 months (end of Term 2, 2010 to end of Term 4, 2011), with a 12-month tapered maintenance period in 2012. Transform-Us! is funded by a National Health and Medical Research Council Grant (No.533815). Ethical approval was obtained from the Deakin University Human Research Ethics Committee (EC 141-2009), the Victorian Department of Education and Early Childhood Development (2009_000344) and the Catholic Education Office (Project Number 1545).