Shown a row of drawn children, 77 of 88 Shanghai parents whose preschooler carried extra weight picked a slimmer body than their child's

Most parents think they would notice if their child were getting heavy. Decades of research say otherwise: parents, especially of young children, tend to see their child as lighter than the growth charts do. What is less clear is whether the answer depends on how you ask. A study published in Nutrients on 25 October put the same parents through two versions of the question, one in words and one in pictures, and compared both with a tape measure and scales.

The work was done between December 2021 and January 2022 in eight kindergartens in Pudong, Shanghai, with researchers from Shanghai Jiao Tong University School of Medicine. Child health staff measured 520 children aged three to six, average age just under five, and the parent who handled their daily meals filled in a questionnaire; three in four were mothers. Using the WHO's BMI-for-age charts, 89 children (about 18%) had overweight or obesity, almost all the rest were in the normal range, and only one was underweight. These were mostly well-off, well-educated families: 86% of parents had been to college.

In words, parents chose from five labels running from "markedly underweight" to "markedly overweight". In pictures, they saw seven age- and sex-matched drawings of children, from very thin to heavy, each tied to a point on the BMI growth charts, and picked the one most like their child. Among the 88 children with excess weight whose parents answered the picture question, 72 parents chose a normal-weight drawing, five chose an underweight one, and only 11 chose one of the two heavier figures. That is 87.5% underestimating. Asked in words, 50 of the 89 parents did call their child overweight, and 43% underestimated.

The underestimation ran the other way too. Around 36% of parents of normal-weight children described or pictured them as underweight, by either method, which matters because worry about a child being too thin often comes with pushing them to eat more. Overall, the words matched the measurements only fairly (weighted kappa 0.37), and the pictures did worse (0.21); counted strictly category by category, the picture answers agreed with the measurements no better than chance.

The authors suggest the two questions tap different things. Saying "overweight" out loud carries stigma, but in China it may be softened by the old idea that a chubby child is a healthy, well-fed one. Choosing a picture is a quicker, gut-level judgement, and when heavier children are common, eyes recalibrate: a heavier body starts to look ordinary. That second idea, called visual normalisation, would explain why the drawings missed so much excess weight.

Did misjudging change how parents fed their children? Parents who verbally underestimated their child's weight scored a little lower on limiting which foods the child ate and on setting rules around eating, but the differences were small and disappeared once the researchers corrected for testing twelve links at once. Picture-based underestimation showed no consistent link with feeding at all. This was a single snapshot from one district, using parents' own reports of their feeding habits, so it shows patterns rather than causes. The practical point is narrower and still useful: the way a doctor or nurse asks about a child's weight changes the answer they get, and a growth chart is a better guide than either a parent's word or their eye.