| Outcome | Mean difference (95% CI) or odds ratio (95% CI) for outcome at 5 months comparing intervention to control group* | p Value |
| Taking account of clustering within schools | ||
| Screen viewing on weekdays (minutes) | −11.6 (−42.7 to 19.4) | 0.42 |
| Screen viewing on Saturdays (minutes) | −15.4 (−57.5 to 26.8) | 0.50 |
| BMI | 0.10 (−0.27 to 0.46) | 0.58 |
| Obesity (odds ratio) | 0.68 (0.24 to 1.94) | 0.68 |
| Walks/cycles to and from school (odds ratio) | 0.45 (0.18 to 1.14) | 0.09 |
| Without taking account of clustering within schools | ||
| Screen viewing on weekdays (minutes) | −12.9 (−38.6 to 12.7) | 0.32 |
| Screen viewing on Saturdays (minutes) | −15.1 (−46.1 to 16.0) | 0.34 |
| BMI | 0.10 (−0.18 to 0.39) | 0.50 |
| Obesity (odds ratio) | 0.68 (0.27 to 1.74) | 0.42 |
| Walks/cycles to and from school (odds ratio) | 0.45 (0.27 to 0.75) | 0.002 |
*All results are adjusted for age, sex and baseline characteristic (eg, screen viewing on weekdays at follow-up is adjusted for screen viewing on weekdays at baseline).
For the mean differences in screen viewing and BMI, the null value is 0; a negative value indicates that the mean level of the outcome is lower in the intervention group compared with the control group (ie, for these three outcomes a negative value indicates a beneficial effect for pupils in the intervention schools); a positive value indicates the mean level of the outcome is higher in the intervention group compared with the control group.
For the odds ratios of obesity and walking/cycling to school the null value is 1; a value below 1 indicates reduced odds of the outcome in the intervention group compared with the control group (ie, for obesity a value below 1 indicates a beneficial effect for pupils in the intervention schools whereas for walking/cycling to school a value below 1 indicates that the pupils from the intervention schools were less likely to walk or cycle to school at the end of the intervention).