Long term effects of treatment of childhood obesity are not well documented but there is growing evidence that parental involvement and behavioral changes are strong predictors of children weight loss. However, which form and content of parental involvement are most effective is not studied. In the present randomized controlled study we compare the effect of parent manualized group treatment ("experimental group") to the effect of parent self-help groups on changes in children Body Mass Index, food intake, physical activity, quality of life and self esteem. We pose the following hypotheses:

1. Parents participating in the experimental group will have children who achieve a larger reduction in BMI than children with their parents in the control group.
2. This treatment effect will be mediated by changes in one of several elements of parents' cognition: outcome expectancies, perceived control, perceived value of outcome, self-efficacy, perceived reduction in barriers, and subjective norms.
3. Reduction in BMI will correlate with increased quality of life, reduced number and severity of mental health problems, and increased self-concept.

The prevalence of obesity as increased dramatically in substantial parts of the world during the last decade, a weight increase also seen in children. This epidemic has also reached Norway, with a worrying increase in the prevalence of overweight and obese school aged children. Reports from US studies show a substantial increase in diabetes II among overweight children. Such an increase is currently not detected among Norwegian children. However, if the increase in childhood obesity continues or worsens, we should expect the onset of weight related disorders such as Diabetes II and Cardio-vascular to take place earlier in life. In addition, there would be an expected increase in psychosocial problems (poor self-concept, social isolation) and psychiatric symptoms and disorders associated with obesity (e.g. eating problems and depression). Hence, both preventive and treatment efforts are called for.

Parents need to be included The treatment of obesity among children has traditionally addressed the child by means of diet and exercise, often involving in-patient treatment. Slowly, the evidence concerning the importance of including the parents in the treatment have emerged as well as the need for addressing the child's wider social context. Hence, childhood obesity should first and foremost be conceived of as a behavioral problem. Generally, behaviorally oriented approaches have emerged as the most successful. Such behavioral procedures have been conducted within a framework of individual consultations to each child/family. It was reported that after family oriented behavioral treatment 30 % of the previously overweight children no longer qualified for overweight, whereas 34 % had sustained minimum 20 % reduction in their overweight. The best predictors of the long term course of overweight reduction were the family eating and exercise environment, in addition to support from friends and family. Notably, the intervention group targeting both children and parents fared best. Such a finding concurs with the conclusion that family and parent based approaches are associated with weight loss among obese children.

Less is known concerning the specifics of effective parental involvement, that is which form of parental involvement are most effective and the content of effective interventions. Therapies targeting individual families are costly to undertake and time consuming for the family. We therefore need to balance the benefits of family treatment with their costs, aiming at an optimal intervention level that provides the best therapeutic results with the lowest degree of investments. We will therefore adopt a group treatment paradigm.

Lessons should be learned from the treatment of other childhood behavioral problems.

Moreover, interventions within pediatric obesity has to a surprisingly little extent drawn upon the mounting knowledge from successful interventions regarding other types of childhood behavioral problems, e.g. conduct problems. Such interventions generally try to strengthen general parenting practices (e.g. limit setting, consistency, anger management, affective availability) in order to increase the parents' competencies in regulating the child's problem behavior. The behavioral techniques that traditional obesity intervention programs try to get parents to adapt at home (e.g. getting their children to exercise more or eat smaller portions) seem to require some basic parenting skills. When parents fall short of actually doing what they have learned, it may be due to the fact that regarding eating and physical activity, the behavior of obese children are especially demanding and putting parenting skills to the test. Hence, specific training of skills that increase the likelihood of behavioral change in the parent and in the child should be included in intervention programs. In doing so, manualized treatment should be encouraged in order to facilitate replications and clinical use, but is often left wanting. Although some exception do exist, many studies only include short-term post-treatment outcomes only. Thus, children need to be followed for longer time periods, due to the fact that short-term weight reduction is achieved by numerous intervention strategies, whereas long-term weight reduction should be the primary goal.

Lessons should be learned from general models of health behavior Within a preventive framework several theoretical models have been devised pinpointing the cognitive and social factors determining health promoting behaviors, also including health behaviors important for the development of adiposity in childhood. The predictive power of such models in determining future health behaviors is high, also when eating and physical activity in children and adolescents are considered. Such basic knowledge has only to a very limited extent been integrated into treatment models of obesity in childhood. We hence aim at targeting those behaviors in parents that may facilitate health promoting behavior in their offspring. Numerous models of health behaviors exist. Their construct overlap to some extent, but each have some merits of their own. Our vantage point has therefore been the arguably most widely accepted model of health behavior, namely Ajzen's Theory of Planned Behavior, and supplementing it with the barriers towards health promoting behavior element stemming from the "Health Belief Model" and Bandura's concept of self-efficacy (1977), viz. the person's beliefs in her/his ability to perform the behavior in question. Figure 1 depicts our theoretical model for the content of the parent based group intervention program. In addition, we expect that weight reduction will imply additional positive outcomes for the child including increased quality of life, increased self-concept and reduction in psychiatric symptoms. However, we should expect that such characteristics of the child may moderate the effect of the intervention.

The considerations above imply that treatment programs for childhood obesity should:

1. Aim for enduring life-style changes with respect to eating and physical activity. The treatment should target the family members' habits and cognitions. Low intensity interventions are the ones that are applicable in most practical contexts, and such programs demand less extensive efforts and changes of the family's way of living, are ones that also have the potential for success in the long run.
2. Target the children's behavior by means of involving their parents. We contend that it is more effective to have the parents regulating their child's eating and exercising behavior than having health personnel trying to do this directly by treating the child on his or her own. Parents can thus regulate the stimulus conditions that govern the child's eating and exercising behavior by reducing cues and possibilities for excessive intake of fattening foods, and increasing cues and opportunities for physical activity and healthy eating.
3. Strengthen general and specific parenting skills. Teaching parents about the importance of healthy eating, meal preparation, and physical activity is not enough to achieve enduring effects. The main challenge is not for parents not knowing what to do, but rather actually being able to carry through the behavior in front of challenges or barriers (e.g. being bereaved of the opportunity to having a snack themselves; guilt-inducing behavior from the child). By addressing their own behavior and by teaching specific skills to overcome such barriers, parents may learn new parenting skills that alter the conditions for their child's behavior in a long-term perspective.
4. Target the parents' cognitions about the effect of their own behavior. The treatment should address those cognitions that maintain their weight related behavior towards the child. These cognitions are outlined in Figure 1.
5. Treat parents in groups. Group treatments are cost-effective. In addition, parents of obese children may learn from the experience of other parents, and they may provide emotional and social support. Finally, the social control element of reporting of completion or failure to perform homework between sessions should not be underestimated. However, in order to secure that the needs of each family/child are met, individual sessions including weight control should be scheduled.
6. Focus on mastery and solutions. In order to promote treatment optimism the treatment should address what the family actually has achieved at the expense of failures.

At present we have little knowledge about the demographics and psychosocial characteristics of clinic-referred children with adiposity in Norway. We will provide descriptive information about this population concerning medical conditions and nutritional status, as well as psychological factors and potential psychiatric symptoms associated with overweight. Moreover, we will address the issue whether successful treatment also will alleviate psychosocial problems such as self-concept problems and increases quality of life. Finally, we ask whether weight reduction in the child are associated with changes in the parents' health beliefs, whether the expected changes in parental health beliefs mediate the effect of the intervention.

Inclusion Criteria:

• Age 7-12 years
• Body Mass Index > 2SDS (age adjusted BMI)
• no known medical cause of obesity

Exclusion Criteria:

• pervasive developmental disorders
• serious psychopathology
• parental drug abuse