Why Better School Food Does Not Automatically Lower Obesity

This chapter examines the apparent contradiction at the heart of school food policy: children’s diets can improve while BMI or obesity measures show little change. The goal is not to dismiss the dietary effects or to promise a hidden weight effect. It is to diagnose where the pathway from school food environment to adiposity becomes longer, noisier and more dependent on conditions outside the cafeteria.

The distinction is essential because “healthier diet” and “lower obesity” describe different levels of the causal chain. One concerns the composition of food consumed; the other concerns a body-size outcome that reflects sustained patterns over time. A school policy can move the first without yet establishing the second.

The endpoint ladder: from composition to adiposity

The evidence can be arranged as a ladder. At the first level, a policy changes what food is available or how it is presented. At the next, children select, purchase and consume some of those options. A broader measure asks what they eat habitually across school and non-school settings. Only after that come outcomes such as BMI, overweight or obesity, and even those are not interchangeable with one another.

This sequence is not a claim that every study must measure every level. It is a way to avoid asking a downstream question of an upstream intervention. If a school standard limits sugary drinks in the cafeteria, the first test is whether availability and purchases changed. If fruit is provided, the immediate test is fruit consumption. A BMI result belongs later in the chain and requires a sufficiently sustained change in the child’s broader exposure.

The PLOS review illustrates the separation. It found improvements in several dietary targets but reported that measures of adiposity were generally unchanged. Direct provision and competitive-food standards did not produce a consistent pooled improvement in BMI, overweight or obesity measures. School-meal-standard studies included some favourable individual findings, including a decrease in BMI percentile in two studies, but other adiposity measures were unchanged. The overall evidence did not support a universal claim that changing school food reliably lowers obesity over the observed periods.

That result is compatible with a diet becoming healthier in composition. A meal can contain more fruit, less sodium or less saturated fat without reducing total calories. A change in nutrient composition may matter for metabolic health or longer-term disease risk even when body size does not shift in the short or medium term. The PLOS authors make this distinction explicitly: the policies may improve diet quality or composition, while obesity outcomes may be more sensitive to dietary quantity over the period studied. Few included studies assessed metabolic outcomes, and those findings were mixed, so even that possible pathway remains less established than the dietary results.

It also helps to separate the different weight-related outcomes that are often compressed into the word “obesity.” BMI is a numerical relation between weight and height, while studies may instead report BMI percentile, the proportion of children above an overweight or obesity threshold, or the odds of crossing such a threshold. A change in one measure does not guarantee a change in another.