Background

Inter-individual differences in adiposity as a result of exercise in overweight and obese children and adolescents are not known. The purpose of this study was to address this gap.

Methods

Using the aggregate data meta-analytic approach, randomized controlled trials (RCTs) on the effects of exercise on body mass index (BMI in kilogram-meters-squared) in overweight and obese children and adolescents were included. Exercise minus control group effect sizes (ES) for BMI from each study were calculated and pooled using the inverse-variance heterogeneity (IVhet) model. Exercise minus control change outcome standard deviations (SD) for BMI from each study were then calculated and pooled using the IVhet model, thus allowing for true inter-individual differences to be examined.

Results

Forty-six RCTs representing 2,290 overweight and obese children and adolescents (1,313 exercise, 977 control) with mean group ages ranging from 8-17 years (mean +/- SD = 12.9 +/- 2.5, Median = 14) were included. Length of training ranged from 6-36 weeks (13 +/- 7, Median = 12), frequency from 1-7 times per week (3 +/- 1, Median = 3), and duration from 6-90 minutes per bout (45 +/- 21, Median = 45). An ES reduction of -0.90 (95% CI, -1.50, -0.32) was found for BMI, the 95% prediction interval (PI) was -3.07 to 1.27, and the probability (percent chance) that the observed change would be clinically meaningful (>1.5 reduction in BMI) in a future study was 30% (possibly clinically important). The pooled SD for true inter-individual responses was -0.16 (95% CI, -0.65 to 0.61), the 95% PI was -2.62 to 2.58, and the probability that the true individual response variability would be clinically meaningful in a future study was 24% (unlikely).

Conclusions

Exercise induced changes in BMI may be clinically important but there is limited evidence to support the belief that there are clinically relevant inter-individual differences in BMI as a result of exercise in overweight and obese children and adolescents.