Table 2.
Overall findings of included meta-analyses.
| Meta-analyses | ES/participants (No.) | Mean (95% CI) | Q (p) | I 2 (%) | PI (95%) |
|---|---|---|---|---|---|
| Atlantis et al. [42] | |||||
| (i) Percent fat (SMD) | 9/369 | −0.4 (−0.7, −0.1) | 15.5 (0.05)* | 48.4 | −1.3, 0.5 |
| (ii) BMI | 3/71 | −1.2 (−3.1, 0.8) | — | — | NA |
| (iii) Body weight (kg) | 11/334 | −2.7 (−6.1, 0.8) | 17.9 (0.07) | 44.1 | NA |
| (iv) Central obesity (SMD) | 4/156 | −0.2 (−0.5, 0.06) | 1.4 (0.70) | 0 | NA |
| McGovern et al. [43] | |||||
| (i) Percent fat (SMD) | 6/358 |
−0.5 (−0.7, −0.3) | 2.2 (0.81) |
0 |
−0.8, −0.2 |
| (ii) BMI-related (SMD) | 11/433 | −0.02 (−0.21, 0.18) | 9.4 (0.49) | 0 | NA |
Notes: No: Number; ES: effect size; CI: confidence intervals; Q: Cochran's Q statistic and associated alpha (p) value for heterogeneity; I 2: I-squared statistic for heterogeneity; PI: prediction intervals, based on a random effects model; SMD: standardized mean difference; central obesity measures derived from waist circumference, waist-to-hip ratio, visceral adipose tissue; BMI-related measures include body mass index (BMI) in kg/m2, BMI z-score, BMI percentile, and percent overweight, adjusted for height. NA: not applicable; —: insufficient data to calculate; mean (95% CI) based on random effects model; boldfaced values indicate continuous data with non-overlapping confidence intervals; *statistically significant, P ≤ 0.05.