Background: The optimal approach to treating childhood obesity remains elusive. The rationale for a family-based approach rests on the fact that obesity tends to cluster in families, and that some developmental stages in children may be particularly suited to family involvement. It is unclear which family-based strategy might help children lose weight. Parents report that they believe they have the nutritional knowledge to guide their children. If this is true, it may be possible that a behavior modification approach could work better than educational counseling alone. Golley and colleagues sought to determine whether the body mass index (BMI) outcomes of prepubertal children could be improved over a 12-month period when parents received combined intensive lifestyle and parenting skills training.

The Study: Overweight Australian children six to nine years of age and their families were recruited through media publicity and newsletters. Children with a BMI z score higher than 3.5 and those with developmental or medical problems were excluded. Children were randomized in single-blind fashion to parenting skills training combined with intensive lifestyle training (P+DA) or parenting skills training alone (P). The control group was composed of families wait-listed for the intervention (WLC).

The P group received workshop and telephone session training adapted to support healthy eating and activity behaviors. The P+DA group underwent the same parenting training but also received seven intensive lifestyle support sessions with a focus on core nutritional knowledge, coping with psychological repercussions of obesity, roles and responsibilities related to eating behaviors, and activity enhancement. The study was carried out over 12 months starting with a baseline demographic survey, BMI measurements that were converted to a United Kingdom-based z score, blood pressure, and laboratory measurements. The primary outcome measure was BMI z score. Waist circumference z scores were also calculated, and parental satisfaction with the interventions was assessed at the end of the study.

Results: Of 262 initially eligible participants, 111 completed baseline measures and 36, 37, and 38 children were allocated to the WLC, P, and P+DA groups, respectively. Twelve-month follow-up was available for 31 children in the WLC group, 29 in the P group, and 31 in the P+DA group. There was a 9 percent reduction (range of −85 to 18 percent) in BMI z score in the P+DA group, a 6 percent reduction (range of −48 to 49 percent) in the P group, and a 5 percent reduction (range of −78 to 16 percent) in the WLC group; these differences were not statistically significant. However, a post hoc group-by-time analysis showed that boys in both intervention groups had significantly greater reductions in BMI z score from baseline compared with the girls. In another group-by-time analysis, waist circumference z scores were significantly lower in the intervention groups than in the control group. There were no differences in metabolic outcomes. Overall parental satisfaction with the program was high.

Discussion: In this study of a family-based program involving parenting skills and lifestyle modification, all groups lost weight, but there were no significant differences in BMI z scores among the three groups. More children in the P+DA group had lower BMI z scores after 12 months compared with the other two groups. Waist circumference z scores were lower in the intervention groups compared with the control group at 12 months, and boys appear to benefit from intervention more than girls. Overall, the benefits of family training and involvement in reducing adiposity were modest. Because the control group also lost weight, the study might not have been adequately powered to detect greater differences. The authors conclude that further studies exploring a sex differential are warranted.

editor's note: A BMI z score reflects each person's BMI relative to the general population. One review observes that reference measurements, such as z score and centile measurements, are adequate for single BMI measurements, but do not assess within-child BMI changes over time.¹ However, if a raw BMI score is used, adult obesity definitions do not apply because BMI is lower in children than in adults.²—c.w.