Table 1. Key Input Parameters and Data Sources in the Dietary Cancer Outcome Model.
| Model input | Outcome | Estimates | Distribution | Comments | Data source |
|---|---|---|---|---|---|
| 1. Simulated population | Population | Mean consumption of added sugars was 52.6 g/d from packaged foods and beverages (eTables 6-8 in the Supplement) | γ | Stratified by age, sex, race/ethnicity; baseline added-sugar intakes were estimated for each of the 32 subgroups | NHANES 2013-2016 |
| 2. Policy impacta | |||||
| Consumer behavior | Policy impact estimate | 6.6% (95% CI, 4.4%-8.8%) | β | A 6.6% reduction in added-sugar consumption from packaged foods and beverages as a result of policy implementation; this was assumed as an 1-time impact | A meta-analysis of labeling interventions on reducing calorie intake7 |
| Industry response | Policy impact estimate | 8.25% (95% CI, 7.5%-9.0%) | β | Assumption: no reformulation in the 1st year of policy intervention; 7.5%-9.0% of the sugar-containing products are reformulated each of years 2-5 of the intervention to achieve a 25% reduction in added sugar content, resulting in a reduction of 8.25% of added-sugar intake associated with the policy intervention | FDA’s Regulatory Impact Analysis; UK sugar reduction strategy8,9 |
| 3. Association between change in added sugar intake (20 g/d) and change in BMIa | Diet-BMI association | Among individuals with BMI<25: 0.10 (95% CI, 0.05-0.15; BMI≥25: 0.23 (95% CI, 0.14-0.32) | Normal | Each 20-g/d reduction in added sugar leads to a 0.1-point reduction in BMI among healthy-weight individuals and a 0.23-point BMI reduction among overweight/obese individuals | A meta-analysis of prospective cohort studies12 |
| Assumption: an 8-oz sugar-sweetened beverage contains 20 g of added sugar based on NHANES; non–sugar-sweetened beverage added sugars has the same impact; the association between added sugar and BMI change would be maintained over a lifetime | |||||
| 4. Association between BMI and cancer risksa | Cancer outcome | RR ranged from 1.05 to 1.50 (eTable 9 in the Supplement) | Log normal | BMI change and cancer incidence | Continuous Update Project conducted by the World Cancer Research Fund/American Institute for Cancer Research13 |
| 5. Cancer statisticsa | Cancer incidence and survival | eAppendixes 1 and 2 in the Supplement | β | Stratified by age, sex, and race/ethnicity | NCI’s Surveillance, Epidemiology, and End Results Program Database; CDC’s National Program of Cancer Registries Database1 |
| 6. Health care–related costsa,b | Medical expenditures, productivity loss, and patient time costs | eTables 11 and 12 in the Supplement | γ | Stratified by age and sex | NCI’s Cancer Prevalence and Cost of Care Projections; published literature14,17,18,19,20,21 |
| 7. Policy costsa,b | For government and industry | eTable 2 in the Supplement | γ | Administration and monitoring costs for government; compliance and reformulation cost for industry | FDA’s budget report; Nutrition Review Project; and FDA’s RIA8,15,16 |
| 8. Health related quality of lifea | For 13 types of cancers | Ranged from 0.64 to 0.86 (eTable 10 in the Supplement) | β | EQ-5D data from published literatures by cancer typec | Published literature18,19,20,21,22,23,24 |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); CDC, Centers for Disease Control and Prevention; FDA, US Food and Drug Administration; NCI, National Cancer Institute; NHANES, National Health and Nutrition Examination Survey; RR, relative risk.
Uncertainty distributions were incorporated in the probabilistic sensitivity analyses. Uncertainties in each parameter are presented in eTable 2 and eTables 6-12 in the Supplement.
If the original source did not provide uncertainty estimates, we assumed the SEs were 20% of the mean estimate to generate γ distribution.
EQ-5D is a standardized instrument developed by the EuroQol Group as a measure of health-related quality of life that can be used in a wide range of health conditions and treatments.