It may not be only how much rice and sugar you eat, but how much it swings from day to day

Most diet advice is about how much carbohydrate we eat and what kind: whole grains rather than refined, fibre rather than sugar. A study published online in Diabetes, Obesity and Metabolism on 29 September 2026, from the School of Public Health at Shanghai Jiao Tong University, looked at something else: how steady that intake is from one day to the next.

The team used UK Biobank, where 53,691 adults (average age 56) recorded everything they ate over 24 hours on at least three separate days. For each person they measured how much their total carbohydrate, fibre and sugar intake swung between those days. Over a median 6.6 years, each step up in variability (one standard deviation) went with 0.20 to 0.24 kg more weight gained. Over a median 10.3 years, more variable fibre intake went with a 5% higher risk of type 2 diabetes and more variable sugar intake with a 9% higher risk. Weight gain explained only a small part of this: about 7% for fibre and 5% for sugar.

These are modest differences, and the fibre result sits right at the edge of statistical significance. The study is observational, so it shows a link, not that irregular eating causes diabetes; people with erratic diets may also differ in sleep, work hours or stress. A few days of food records is also a rough picture of anyone's habits. The authors' conclusion is careful: regular carbohydrate intake may be one more part of a healthy diet worth studying, alongside the amount and quality.