How a simple questionnaire about children’s eating behaviour could help match them to the right kind of obesity support

Childhood obesity is becoming more common across the world, and treating it is rarely straightforward. Different children have very different relationships with food, which means the same programme won’t work equally well for everyone. To make treatment more personalised, researchers have increasingly turned to children’s individual appetitive traits — the tendencies that describe how a child eats.

Some children, for example, are highly food responsive: they’re easily drawn in by food cues around them, such as seeing others eat or spotting an advert for a snack they like. Others have low satiety responsiveness: they do not notice inner feelings of fullness and miss bodily signals to stop eating. Neither is “good” or “bad” — but they point towards different kinds of support.

Measuring how children eat

These traits can be measured with the Child Eating Behaviour Questionnaire (CEBQ), a widely used parent-report tool that captures eight appetitive traits in total. Four are food-approach traits, reflecting a keener appetite — Food Responsiveness, Enjoyment of Food, Emotional Over-eating and Desire to Drink. The other four are food-avoidant traits, reflecting a less avid appetite — Satiety Responsiveness, Slowness in Eating, Emotional Under-eating and Food Fussiness.

New treatments that target these traits directly — for instance, training children to notice their own hunger and fullness signals before eating — have shown real promise. But that success surfaces a practical problem: if different children need different treatments, we need a reliable way to work out which child should be offered which kind of support. Until now, that tool has been missing.

What we did

A team of researchers based at IMS Epidemiology, University of Cambridge, UCL, and the University of Leeds set out to test whether the eight CEBQ traits could serve as that tool — and, crucially, what score a child would need in order to be flagged for a particular targeted treatment. To do this, they pooled 17 international datasets covering 15,273 children aged 2 to 18, then tested how well each trait distinguished children with obesity from those without, and derived cut-off scores above which a child was likely to have obesity based on that trait. A cut-off score is the threshold used to decide whether someone falls into a particular category: It is not always the same as a diagnosis, but it provides a practical decision rule based on data.

What we found

Six of the eight traits proved useful — Food Responsiveness, Enjoyment of Food, Emotional Over-eating, Desire to Drink, Satiety Responsiveness and Slowness in Eating — though to varying degrees and only in some age and sex groups. Two patterns stood out. The traits worked better in middle childhood (roughly 7 to 12 years) than in younger children, possibly because obesity is rarer very early in life and because older, more independent children show clearer differences in how they eat. And Food Responsiveness performed best overall: among girls aged 7 to 12, a score above 2.6 correctly identified around 8 in 10 of the children who had obesity, and about the same proportion of those who did not.

In practice, a score like that could offer a child a treatment built around food responsiveness — say, learning to manage reactions to food cues — while another child is pointed towards support that better fits their eating behaviour.

Important caveats

These results come with some caveats, and we want to be upfront about them. How well the cut-offs worked varied considerably from study to study, so future research needs to unpick what drives this variation before these scores can be relied on in everyday practice. We also stress that any such score should be used sensitively: as one part of a broader clinical assessment rather than a label, ideally alongside other measurements, and always communicated to families in a way that avoids stigma.

Why it matters

This is the first study to derive CEBQ score thresholds that can be used to characterise a child’s obesity through their eating behaviour. Used carefully, and always alongside professional judgement, these scores could help match children to the support most likely to help them — both in research, where they could help assign children to studies testing specific treatments, and in clinical practice, where they could inform decisions about which children might benefit from which kind of behavioural support.

Lead author, Dr Laura Kudlek said:

The CEBQ is easily accessible for clinics and research studies around the world, so a set of cut-off scores gives clinicians a practical, low-cost way to look beyond weight alone and understand how a child eats — and to tailor support to that child rather than offering everyone the same programme.”

Laura also explained:

For too long, obesity treatment has been one-size-fits-all. If we can reliably identify a child’s eating-behaviour profile, we open the door to genuinely personalised behavioural weight management — the right support for the right child at the right time. This study is a first step towards making that possible.”

This post is based on the open-access paper: Kudlek L, Smith AD, Ruggiero CF, et al. “Identifying eating behaviour phenotype cut-off scores for childhood obesity: an individual participant data meta-analysis using the Child Eating Behaviour Questionnaire (CEBQ).” International Journal of Behavioral Nutrition and Physical Activity (2026). Read the full paper here: https://doi.org/10.1186/s12966-026-01951-8