Fig. 2

Treatment effect over differing levels of baseline sedentary time. Figure 3 shows the point estimate of the treatment effect (difference in mean weekday MVPA at T1) as a linear function of the proportion of sedentary time reported at T0 along with the 95% confidence interval bounds. We are not powered to draw conclusions from these estimates or confidence intervals, but this analysis could suggest that the Plan-A intervention has a more positive effect on those who have higher proportions of sedentary time prior to the intervention.