Childhood obesity is one of the most prevalent health issues affecting children in developed countries globally. The International Obesity Task Force  reported that there are more than 155 million children worldwide who are overweight or obese. Over the past 3 decades, the rate of adolescent obesity has tripled in Canada, from 3% in 1978 to 9% in 2004 . The significance of this emerging trend is the connection between childhood obesity and type 2 diabetes (T2D), cardiovascular disease (CVD), psychosocial problems (i.e., depression and low self-esteem) and adult obesity [3–5]. A reduction in obesity and associated risk factors may prevent T2D and CVD in this high-risk pediatric population [4, 6]. Two major strategies for management of obesity and associated metabolic abnormalities are lifestyle modification  and pharmacologic therapy .
Wilfley and colleagues  conducted a meta-analytic review of RCTs for the treatment of childhood obesity and concluded that lifestyle interventions were effective at decreasing participants' body mass indexes (BMI) in the short-term and reported some evidence for long-term positive effects. Interestingly, Wilfley and colleagues  found that wait-list or control group participants increased in percentage overweight from 2.1% (immediately following treatment) to 2.7% (at follow-up). Thus, without treatment, it can be expected that the average participant will continue to gain weight. These findings also suggest that treatment programs that result in slight improvements or maintenance of participant's percentage overweight should be utilized.
Although positive findings were reported for improvements in percentage overweight, Wilfley and colleagues point out a number of limitations to the RCTs reviewed and make several recommendations for future research. Their first recommendation is that treatment programs implement sufficient long-term follow-up assessments, occurring at least 1-year and ideally 2-years post-randomization. Longer-term follow-ups allow for the determination of factors that protect and/or are associated with diminished treatment effects over time. Wilfley and colleagues  also suggest that in order to accurately assess changes in participants' percentage overweight, childhood obesity treatment studies should utilize weight outcome assessments that take into account changes in height. Reporting results that highlight the clinical importance of the findings, including comprehensive participant characteristics, analyses used, and other measures of health and psychosocial functioning (i.e., effects on psychological factors and comorbidities, such as T2D) is also recommended . Finally, they suggest that future studies elucidate variables that moderate and/or mediate the long-term effectiveness of childhood obesity treatment programs [i.e., environmental or psychological factors; ].
Importance of Theory-based Interventions
In developing an intervention aimed at changing individuals' behaviour, it is important to use behaviour change theories and models to guide the intervention. Based on established principles, theory-based interventions aid in the planning, implementing and evaluating stages of an intervention . Furthermore, theory-based interventions allow researchers to make predictions about why certain changes occur and help determine potential mechanisms through which the intervention is working . For example, theories of behaviour change suggest that psychosocial factors play an important role in changing behaviour. In addition, theories of behaviour change suggest important variables to measure and methods of assessing the intervention. In doing so, researchers are able to identify specific aspects of the intervention that worked, and aspects that need improvement.
One of the most influential theories of health behaviour change is social cognitive theory [SCT; ]. SCT describes human behaviour as an interaction between the individual, behaviour and environment and has been successfully used to understand, predict and change behaviour . Therefore, an SCT-based childhood obesity intervention would focus on individuals and their behaviour, as well as the environment (i.e., the individual's family and their living situation) in order to help them make healthy lifestyle changes. A model that incorporates these variables is the family-based multidisciplinary obesity treatment model . This model was developed over 25 years ago and has demonstrated both short- and long-term efficacy . Families are generally referred to these programs based on one (or more) of the children's health status (i.e., high blood lipid levels or high BMI). Typically, multidisciplinary family-based treatment programs involve a psychologist or social worker, a pediatrician, a dietitian and a physical trainer. In line with SCT, the health care providers teach the families ways to change their behaviours for their particular living situation. For example, behaviour modification techniques taught in such programs incorporate self-regulation skills, including goal setting, self-monitoring, and corrective behaviours (e.g., including three of four food groups in each meal, planning for regular physical activity).
Multidisciplinary family-based childhood obesity treatment programs are able to address a number of the limitations outlined in the Wilfley and colleagues review. For example, such treatment programs are structured to involve long-term treatment and follow-up. As previously mentioned, these programs involve specialized health care providers (i.e., behavioural, physical, nutritional, psychological and medical), which allows for more thorough assessments to be administered (i.e., markers of T2D and CVD, detailed dietary reviews, objective assessments of physical activity). Furthermore, the interpretation and application of results can be more detailed and clinical significance can be disseminated because all healthcare providers are experts in their field. The specific components of multidisciplinary childhood obesity treatment programs will be discussed.
Behaviour Change Skills Sessions
One of the main goals of a childhood obesity treatment program is to help participants lose weight. Research has demonstrated that obesity interventions have been able to achieve weight loss in the short term [7, 12, 13], and there is some evidence to suggest they achieve long-term weight loss as well . In order for participants to sustain their weight loss, it is essential for obesity treatment programs to teach participants skills to maintain their behaviour changes. The underlying skill in making and maintaining such lifestyle changes is the ability to self-regulate health behaviours, which is the ability to consciously make healthy choices [i.e., engage in regular physical activity and healthy eating; ]. According to SCT theory, self-regulatory skills play an important role in making and maintaining behavioural changes [14, 15]. Self-regulation involves monitoring one's behaviour, comparing one's behaviour to set goals or criteria (evaluation/feedback) and changing the behaviour to meet the set goals . A recent study by Alm and colleagues  found that overweight adolescents who weighed themselves regularly (self-monitoring) engaged in more healthy weight control behaviours (self-regulation), including decreasing their caloric intake, and eating less fatty foods and junk foods. However, effectively teaching these skills can be difficult. Recently, Brawley, Rejeski and Lutes  developed the group-mediated cognitive-behavioural (GMCB) model, which aims to facilitate learning of self-regulation skills through using the group as an agent of change. The GMCB model involves teaching self-regulatory skills throughout the sessions, and increasingly relies on independent self-regulation to teach participants to be independently active.
GMCB interventions are based on SCT  and the group dynamics literature . SCT is used as the theoretical framework because it suggests that to increase adherence to an optional behaviour it is essential to change an individual's cognitions. For example, if the individual is to change their behaviour they must value the outcome of the behaviour, believe they can produce the desired outcome, and believe that the outcome will result from successfully completing the behaviour . Group dynamics theory is used because group cohesion - a central component of group dynamics, is associated with greater adherence to physical activity . Thus, developing a cohesive group is thought to be one of the underlying mechanisms to motivate and enhance learning of behaviour change skills (i.e., self-regulation and goal setting).
Recently, separate studies involving post-natal women  and elderly adults [17, 22, 23] implemented GMCB interventions, and found a positive impact on adherence to an exercise program. These GMCB interventions lead to greater improvements in frequency of exercise [21–23], long-term adherence [21–23], fitness , self-efficacy for mobility , and barrier self-efficacy , as compared to those in control groups. Collectively, GMCB studies are thought to be effective because they facilitate learning self-regulatory skills through a cohesive and supportive group environment, which encourages participants to make and maintain healthy lifestyle changes.
One of the most frequently cited factors contributing to the current childhood obesity epidemic is the high rate of physical inactivity in youth today [24, 25]. The Public Health Agency of Canada  suggests that youth should accumulate at least 90 minutes of moderate and vigorous physical activity per day, and decrease the amount of time spent being sedentary (i.e., watching TV) by at least 90 minutes per day. Disappointingly, based on more conservative guidelines of only 30 minutes of physical activity per day, Sithole and Veugelers  found that only 48.7% of Canadian children are active at least at a moderate intensity. Thus, one of the ways to address the childhood obesity epidemic is to get obese children more physically active. Multidisciplinary obesity treatment programs often involve a specific exercise component, ranging from telling people to start exercising, to having people exercise while at the program [11, 12]. van Sulijs and colleague's  review found that more intensive physical activity interventions were associated with greater improvements in physical activity levels. Further, it is suggested that programs that involve structured physical activity sessions are more likely to be more intensive. Thus, including physical activity sessions in the program may aid in the success of the treatment program.
Nutrition also plays an important role in the development of childhood obesity. Indeed, consuming a diet rich in foods that are lower in nutrient density and fiber and higher in fat and calories, contributes to excessive weight gain . According to the Ontario Ministry of Health and Long Term Care , only 42% of individuals age 12 years and older reported consuming the recommended five or more servings of fruits and vegetables per day. Diets that include a variety of and considerable amount of fruit and vegetable servings are associated with healthy weights, a decreased risk of obesity as well as a decreased risk of other diseases such as cancer and cardiovascular disease [30–32]. Accordingly, nutrition counseling that aims to improve eating habits (i.e. increase fruit and vegetable servings, limit lower nutrient density and higher fat and calorie foods) is a critical component of a successful obesity treatment program .
Given the necessary influence of parents throughout the development of their children, it is essential to involve parents in the treatment of childhood obesity . Based on Bandura's SCT and Epstein's model of family-based interventions, parental modeling (i.e., parents engaging in regular healthy eating and exercise) and parental reinforcement (i.e., parents supporting children in their healthy changes), play an essential role in their child's weight loss. Thus, multidisciplinary treatment programs often involve a family coach (i.e., a social worker or psychologist), who works with the family to address family issues and help parents set a good example for their children in making healthy changes.
In addition to involving the parents, the family coach addresses important issues the child is facing. Childhood obesity research often explores issues of diminished self-esteem and body image, and how they relate to increased weight. Research suggests a strong association between lowered self-esteem and higher BMI measures for elementary school-aged children . Furthermore, research suggests that diminished self-esteem is associated with significant increases in sadness, loneliness and nervousness . These issues are addressed during social work sessions, as early adolescence is a critical period for the development of self-esteem among obese children .
For obese children and adolescents, lifestyle changes alone may have a positive impact. However, the sustainability of these lifestyle changes remains to be evaluated. Preliminary data indicate that the addition of metformin to a lifestyle intervention program is associated with greater reduction in BMI as well as improved modification of metabolic risk factors for T2D and CVD . Metformin often promotes weight loss and is clinically effective in reducing insulin resistance as well as positively impacting lipid levels. The efficacy and safety of metformin in improving BMI and associated metabolic risk factors in obese, non-diabetic adolescents has been demonstrated in several pediatric trials, including a recently published 6 month intervention by Clarson and colleagues , in doses up to 2000 mg daily with duration of therapy from 8-26 weeks [36–38]. Clarson and colleagues  found that BMI and dyslipidemic profiles improved most in participants taking metformin. However, independent of metformin, lifestyle intervention resulted in improvement in participants' metabolic risk factors (i.e., plasma lipids and adipocytokines).
Although these studies all demonstrated improvement in anthropometry in response to up to 6-months of metformin therapy, none were maintained for longer than 6 months. Data on the long-term efficacy of metformin intervention programs for pediatric obesity are lacking, in particular relating to the sustainability of lifestyle interventions.
REACH is a 2-year program including a metformin/placebo intervention, an intensive 12-week exercise program followed by a weekly 1.75 year long-term exercise program, behaviour change techniques, family sessions with a dietitian and a social worker, and comprehensive medical monitoring. The primary outcome measure is change in BMI. The secondary outcome measures are changes in body composition, risk factors for T2D and CVD, diet, PA and psychosocial functioning. It is hypothesized that participants who take metformin and engage in vigorous intensity exercise will show the greatest improvements in body mass index. In addition, it is hypothesized that participants who adhere to the REACH program will show improvements in body composition, physical activity, diet, psychosocial functioning and risk factor profiles for type 2 diabetes and cardiovascular disease. These improvements are expected to be maintained over the 2-year program.