Objective: To investigate associations between weight-status and body-size perception in children in the UK.
Design: Cross-sectional survey.
Setting: School based sample in the UK.
Participants: 399 children (205 boys, 194 girls) aged 7 to 9 years.
Main outcome measures: Body mass index (converted to BMI SD scores using UK data), weight-status and demographic information. Perceived body-size was assessed using a visual method (Children’s Body Image Scale, matching to images representing BMIs from 3rd to 97th percentiles) and verbal descriptors from ‘too thin’ to ‘too fat’.
Results: Modest associations between actual and perceived body-size were found with visual (r=0.43, p<0.001) and verbal (r=0.41, p<0.001) methods, but there was a consistent response bias towards under-estimation. Using visual matching, most children (45%) under-estimated their body-size, with significantly greater under-estimation (p<0.001) at higher BMIs. A gender-by-weight group interaction (p=0.001) showed that at lower weight girls were more accurate than boys, but at higher weights girls were less accurate. Using the verbal scale, the majority of children reported their body-size as ‘just right’ in all weight groups (52-73%), with no sex differences.
Conclusions: Children can estimate their body-size using visual or verbal methods with some accuracy, but show greater under-estimation at higher weights, especially in girls. These findings suggest that under-estimation is more widespread than has been assumed, which has implications for health education among school-aged children.