Since 2015, the UK government has worked with the food industry to reformulate a wide range of food products to reduce sugar, fat and calorie content. The industry has been given the target of lowering the number of calories in certain products by 20% by 2025. The reformulation scheme was overseen by Public Health England (PHE) until 2021, and is now overseen by the Office for Health Improvement and Disparities. The scheme is voluntary, but some organisations have called for it to be mandatory. In modelling published in 2018, PHE acknowledged that lower calorie intake could have a 'potentially negative impact' on people who are a healthy weight or underweight, but it excluded these people from its model. Since being underweight is associated with a number of serious health problems, this was a major omission which we address in this paper by modelling the impact of the calorie reduction scheme on the prevalence of underweight among children. Using two different estimates of baseline energy flux, our model shows that among 4-5-year-olds, the calorie reduction scheme would lead to a ~4% reduction in energy intake, and a reduction in obesity rates between 0.8 and 1.3 percentage points. However, the prevalence of underweight would increase by between 3.0 and 4.8 percentage points. Among 10-11-year-olds, the model shows that energy intake would decline by between 2.6% and 5.4%, leading to a reduction in obesity prevalence of between 0.2 and 1.1 percentage points, but the prevalence of underweight would rise by between 2.1 and 4.1 percentage points. Under every scenario, for each child who moves from the obese category to the healthy weight category, at least two children become underweight. If the reformulation scheme works as intended by reducing calorie intake across the whole population, it will increase the number of underweight 10-11-year-olds by at least 30% and possibly by as much as 60%. This will lead to a net increase in the number of children who are an unhealthy weight. However, neither our model nor PHE's model should be taken seriously as a prediction of what would happen if the targets were met. It is much more likely that consumers would compensate by buying more food (or buying different types of food) to obtain the same number of calories. If so, it will increase the cost of feeding a typical household by around 10%. This compensatory behaviour makes it less likely that significant numbers of children would become underweight as a result of reformulation, but it also makes it less likely that the scheme would have its intended effect of reducing rates of obesity. The surprising results reported in this paper should be taken as an illustration of the flaws in the reformulation theory.
