Table 2. Description of sensitivity analyses.

Sensitivity analysis	Description
2018 nutrient profiling model	We used the revised version of the nutrient profiling model, published in 2018. This model has revised thresholds for sugar and is more liable to classify food with sugar as HFSS [52].
Classification of brand adverts—all HFSS	We assumed all adverts for brands that contained a mix of HFSS and non-HFSS products included HFSS products within the advert and would be subject to the restrictions.
Classification of brand adverts—all non-HFSS	We assumed all adverts for brands that contained a mix of HFSS and non-HFSS products included only non-HFSS products within the advert and would not be subject to the restrictions.
NICE discount rate	We used a discount rate of 3.5% for costs and DALYs (with no tapering over time), instead of those recommended by the UK Treasury. The 3.5% discount rate is recommended by NICE for evaluation of health and care interventions in the UK.
Differential impact by weight	Assuming the same average change in energy intake, we assumed a greater effect of the proposed regulations on children with overweight than on children with normal weight. Based on the Russell et al. [19] meta-analysis, we assumed that children with overweight consumed more calories in response to HFSS advertising than children with normal weight (125.5 kcal versus 79.9 kcal, ratio 1.57); that children with overweight watched more television, and thus were more exposed to HFSS adverts (based on Health Survey for England 2008 data, obese children watch 120 minutes versus 102.9 minutes for those who are not obese; ratio 1.17); and that the cumulative effect of the increased exposure and increased sensitivity to exposure resulted in an approximate 80% difference in change in energy intake (1.57 × 1.17 = 1.83).

DALY, disability-adjusted life year; HFSS, high in fat, sugar, and salt; NICE, National Institute for Health and Care Excellence.