Skip to main content

Table 5 Paired T-tests for BMI SDS, at the child and cluster/school levels, for six measurement comparisons using a Holm adjustment for multiple tests

From: Testing a multicomponent lifestyle intervention for combatting childhood obesity

Paired T-tests (Child level – Primary analysis) Mean/Estimatea SD/SEa 95% CI t-stat p-value
Baseline to Post-Camp -0.09 0.13 (−0.11,-0.08) −11.30 < 0.0001
Baseline to Post-Club − 0.14 0.26 (− 0.17,-0.11) −9.34 < 0.0001
Post-Camp to Post-Club −0.06 0.25 (−0.09,-0.03) −3.79 < 0.0001
Paired T-tests (School level – Sensitivity analysis)
 Baseline to Post-Camp −0.09 0.07 (−0.13,-0.05) −5.24 < 0.0001
 Baseline to Post-Club −0.15 0.12 (−0.21,-0.08) −4.78 0.0002
 Post-Camp to Post-Club −0.07 0.11 (−0.13,-0.002) −2.23 0.0220
Random Effects Models (Sensitivity analysis)
 Baseline to Post-Camp
  alpha −0.09 0.02 (−0.13,-0.06) −5.15 < 0.0001
  sigma 0.06   (0.04,0.09)   
 Baseline to Post-Club
  alpha −0.15 0.03 (− 0.21,-0.08) −4.82 < 0.0001
  sigma 0.10   (0.05,0.16)   
 Post-Camp to Post-Club
  alpha −0.06 0.03 (−0.12,-0.01) −2.32 0.0102
  sigma 0.08   (0.03,0.14)   
  1. aMeans and standard deviations (SDs) correspond to t-tests, while model estimates and standard errors (SEs) correspond to random effects models