Page 61 - Go4it

measures (including body composition measured by DXA and metabolic components). The main limitation of the present trial is the high level of attrition, which was 44% (49% in the intervention group and 37% in the control group) at 18 months. Ball et al. recently showed that even on the short term an attrition rate of 20-40% was found in this target group (27). Also, Nguyen et al. reported 38%, and Savoye et al. even 56% attrition (18,26). Therefore, this level of attrition is comparable to other studies concerning obesity treatment in adolescents (12-47%) (18,21-28). Data imputation was not performed because subjects were in their mid-to-late puberty; therefore increases in height and weight were expected. By imputation of data, a somewhat misleading estimate of the intervention effect would be provided. Another limitation is the asymmetrical randomization. The main reason for this asymmetrical randomization was that to start the group sessions as soon as possible after the first included adolescent was enrolled into the study. Otherwise, it would have taken several months before a sufficient number of adolescents had enrolled to start the group sessions. Disappointing was the lack of a significant treatment effect in non-western obese adolescents, despite tailoring the nutrition education and advice of Go4it to the different ethnicities. Potential explanations are that parents from Turkish or Moroccan descent underestimate the actual weight status of their children as well as the resulting health effects. Thus, cultural norms of healthy body weight and attractive appearance are probable causes of the observed difference (29). Savoye et al. also studied a culturally diverse group of children but did not report on effect-modification (18).
Beforehand, it was intended to develop an intervention that is implementable into the child health care and primary health care setting. This affected choices in the design of the intervention, which had to be evidence based and effective but also practical and affordable. For instance, there was no exercise program provided, but participants were encouraged to seek and participate in existing exercise programs in their neighbourhood. The intervention could have been upgraded with telephone coaching, short-message-service, and email communication at low cost. However, Nguyen et al. showed that this additional therapeutic contact did not improve the outcome for 24 months (30). Since travel distance and the limited time of working parents and schoolchildren was one of the main reasons for non-compliance, implementation of the intervention nearby, home or school may improve compliance.
Conclusions and implications
In summary, the Go4it intervention showed a significant long-term effect on BMIsds, compared to current regular care in obese adolescents. Implementation of Go4it in a setting nearby home or school may even improve successful treatment of obese adolescents. In addition, development of additional programs aimed at the non-western ethnic group is urgently required.
Effectiveness of Go4it
61