A new economic model estimates that England’s calorie-labelling rules for large food businesses could produce a small lifetime health gain at a low public-sector cost, but the result is not a measured outcome. Its central estimate is a saving of £9.15 per person over a lifetime and a gain of 0.0021 quality-adjusted life years. Both estimates carry substantial uncertainty, and the authors say their conclusion depends on modelling assumptions.
What the study calculated
The policy took effect in England on 6 April 2022. It generally requires businesses with at least 250 employees to show calorie information for non-prepacked food and drink sold for immediate consumption. The rules cover a mix of restaurants, cafes, takeaways and online menus, subject to defined exemptions. They do not require businesses to reduce the calories in their dishes.
The research team built a simulated population of people aged 13 to 79 using data from the National Diet and Nutrition Survey and Health Survey for England. Rather than following those individuals through decades of observed outcomes, the model applied short-term estimates of how labelling was associated with consumer choices and menu changes, then projected possible effects across a lifetime.
It translated changes in daily calorie intake into weight trajectories and estimated later effects on obesity, diabetes and complications, cardiovascular events, osteoarthritis, cancers and dementia. The calculations also included quality-adjusted life years, a measure combining length and quality of life, and costs to the NHS and wider public sector. The analysis compared the policy with a hypothetical scenario in which it did not exist.
Small central estimates, wide uncertainty
The central estimate of £9.15 in lifetime savings per person is modest. More importantly, the study’s 95 per cent uncertainty interval runs from £31.63 in savings to £2.90 in additional cost. In other words, the model does not rule out a small net cost. Its estimated gain of 0.0021 quality-adjusted life years per person also has an interval that includes no gain and a possible small loss.
The authors report that 94 per cent of the model’s probabilistic simulations fell below a £20,000 cost-per-quality-adjusted-life-year threshold, commonly used in health-economic assessments. That result supports the description “likely cost-effective” within this analysis. It does not mean that 94 per cent of people will benefit, nor that the policy has been shown to save a definite amount for every person.
For the most deprived group, the model estimated slightly larger savings and health gains than for the population overall: £10.62 saved per person and 0.0022 quality-adjusted life years. The uncertainty intervals again included the possibility of higher costs or no health gain. The finding therefore suggests a possible modest equity benefit, rather than demonstrating that health inequalities have narrowed in practice.
Why this remains a projection
The analysis depends on inputs drawn from real-world evaluations, but it must extend short-term changes into long-term disease and cost outcomes. Estimates can shift if consumer responses fade or grow, if menus change differently from the assumed pattern, or if underlying eating and health trends alter. The authors explicitly warn that their uncertainty calculations do not remove all risks created by model structure or assumptions.
Earlier England modelling likewise found only modest projected effects under the current rules, while estimating larger impacts in a hypothetical extension to every out-of-home business. Those findings are not the same as measured reductions in illness: they are projections based on their own assumptions. Existing research has also examined menu changes and purchasing, which are different outcomes from lifetime disease prevention.
That distinction matters because calorie labels are information, not a requirement to reformulate food or make lower-calorie options available. The policy covers large businesses, leaving many smaller outlets outside its direct scope. Even where information is displayed, a modelled benefit depends on how consumers and businesses respond and on whether those changes persist.
The study does not establish that the policy has already prevented cases of disease, quantify every possible unintended effect, or decide whether the rules should be expanded. Its contribution is a long-term economic estimate built from limited short-term evidence. The most defensible reading is cautiously positive but conditional: the policy may offer small value for public spending, while both the size and certainty of its health benefits remain unresolved.
