Fig. 4

Illustration of difference in biological risk factor development between intervention/control schools in the CHAMPS study-DK (2008, 2010, and 2015). Estimates are mean difference between intervention and control school with a 95% CI. Negative values favour intervention group. Two-year follow-up estimates are from an earlier publication with 712 participants analysed [24]. 6.5 year follow-up are estimates from the 312 participants in this manuscript. Confidence intervals not overlapping zero are statistically significant. Baseline group-differences set to zero as results are presented as adjusted difference in change including baseline-value of outcome, which provides the interpretation that group-mean values are identical at baseline