Study objective. To evaluate the effect of family therapy on childhood obesity. Design. Clinical trial. One year follow-up. Setting. Referral from school after screening. Participants. Of 1774 children (aged 10 to 11), screened for obesity, 44 obese children were divided into two treatment groups. In an untreated control group of 50 obese children, screened in the same manner, body mass index (BMI) values were recorded twice, at 10 to 11 and at 14 years of age. Intervention. Both treatment groups received comparable dietary counseling and medical checkups for a period of 14 to 18 months, while one of the groups also received family therapy. Results. At the 1-year follow-up, when the children were 14 years of age, intention-to-treat analyses were made of the weight and height data for 39 of 44 children in the two treatment groups and for 48 of the 50 control children. The increase of BMI in the family therapy group was less than in the conventional treatment group at the end of treatment, and less than in the control group (P = .04 and P = .02, respectively). Moreover, mean BMI was significantly lower in the family therapy group than in the control group (P < .05), and the family therapy group also had fewer children with BMI > 30 than the control group (P = .02). The reduction of triceps, subscapular, and suprailiac skinfold thicknesses, expressed as percentages of the initial values, was significantly greater in the family therapy group than in the conventional treatment group (P = .03, P = .005 and P = .002, respectively), and their physical fitness was significantly better (P < .05). Conclusions. Family therapy seems to be effective in preventing progression to severe obesity during adolescence if the treatment starts at 10 to 11 years of age.

Obesity in children is difficult to treat, but it seems to be easier to treat than adult obesity. The first step in treatment is to identify effective advice relating to nutrition and physical activity. In most treatment studies the macronutrient composition of the diet is not of major importance for treatment outcome. In relation to physical activity fat-utilisation strategies have been described. The second step includes appropriate approaches to lifestyle change. In Europe there are no drugs approved for children, and surgery for children is still limited to research projects. Thus, the major challenge is to develop effective ways of changing lifestyle. Family therapy may be an effective approach in preventing severe obesity from developing during puberty, and a therapeutic strategy based on treatment studies is described. The family-therapy techniques used here are intended to facilitate the family's own attempts to modify their lifestyle, and to increase their own sense of responsibility and readiness to change, i.e. these variables are the prime targets during therapy. Thus, the family, not the therapist, assumes responsibility for the changes achieved. This approach may be helpful in making the therapeutic process less cumbersome for the therapist. Instead of the therapist attempting to persuade the obese subjects to lose weight, it might be more effective to teach them to control their eating patterns through their own efforts. The treatment model includes structural family therapy and solutionfocused-brief therapy. The use of such a model makes it possible to train therapists and health professionals to use an evidence-based intervention model. Family therapy: Childhood obesity: Systemic family medicineTreating obesity in children, although difficult, seems to be easier than treating adult obesity (1)(2)(3) . The first step is to identify effective advice on nutrition and physical activity.In relation to nutrition, several studies have investigated an epidemiological association between BMI and food habits, but the results have been inconclusive, as pointed out in a recent review (4) . From the epidemiological perspective, the cross-sectional Bogalusa Study, which has investigated 1562 children aged 10 years over a 21-year period, has found that the consumption of sweetened beverages, sweets and meat and the total consumption of low-quality foods are positively associated with overweight status (5) . Furthermore, the total amount of food consumed, specifically from snacks, is positively associated with overweight status. However, the percentage variance explained from the eating pattern-overweight models is very small. In relation to treatment, a meta-analysis has shown that low-carbohydrate non-energy-restricted diets in adults appear to be at least as effective as low-fat energyrestricted diets in inducing weight loss for £ 1 year (6) . However, in a recent study of forty-one 8-12-year-old obese children who were randomly assigned to a 24-month family-based behavioural treatment the macronutrient cont...