Table 3: TP results.

VariableStudies (#) 𝑔 (95% CI)Q (p) 𝐼 2 (95% CI)

PP Analysis
 Overall8−0.68 (−1.16, −0.20)*28.6 (<0.0001)**75.5% (50.9%, 87.8%)
 Type of training
   Aerobic4−0.44 (−1.04, 0.16)11.2 (0.01)**73.2% (24.7%, 50.5%)
  Strength2−1.13 (−1.73, −0.53)*0.5 (0.47)0%
  Both2−0.75 (−2.22, 0.71)10.8 (0.001)**90.7% (66.7%, 97.4%)
 Sedentary prior to enrollment
  Yes5−0.67 (−1.23, −0.08)*16.6 (0.002)**76.0% (41.2%, 90.2%)
  Unclear3−0.71 (−1.73, 0.31)11.8 (0.003)**83.1% (48.6%, 94.4%)

ITT Analysis
 Overall4−0.24 (−0.62, 0.15)10.1 (0.04)**60.6% (0%, 85.2%)
 Type of training
  Aerobic4−0.24 (−0.71, 0.22)10.1 (0.02)**70.3% (37.4%, 93.7%)
  Strength10.25 (−0.96, 0.46)
  Both
 Sedentary prior to enrollment
  Yes2−0.14 (−0.47, 0.20)0.62 (0.43)0%
  Unclear3−0.33 (−1.1, 0.42)9.4 (0.009)**78.8% (32.2%, 93.4%)

Notes: PP, per-protocol; ITT, intention-to-treat; 𝑔 (95% CI), Hedge’s 𝑔 and 95% confidence intervals; Q (p), heterogeneity statistic and alpha value; 𝐼 2 (95% CI), percent inconsistency and 95% confidence interval; —, insufficient data to calculate; *statistically significant within-group difference because 95% confidence intervals do not include 0; **statistically significant heterogeneity (p ≤ 0.10).