Do children really eat what they like? Relationships between liking and intake across laboratory test-meals

Liking plays a primary role in determining what and how much children eat. Despite this, the relationship between liking and intake of foods and beverages served as part of a meal is not often reported, even though pediatric feeding studies frequently collect such data. In addition, few studies have reported on the test-retest reliability of both hedonic ratings and laboratory intake among children. To address these gaps, this study was designed to assess the relationship between children's liking of items at a meal and subsequent intake. 61, 4-6 year-olds were recruited to participate in two identical laboratory sessions where liking of 7 foods (i.e., chicken nuggets, ketchup, potato chips, grapes, broccoli, cherry tomatoes, cookie) and 2 beverages (i.e., fruit punch, milk) was assessed (5-point hedonic scale) prior to ad libitum consumption of the same items at a meal. Spearman's correlations tested the relationship between liking and intake and intra-class correlations assessed intersession reliability of both measures. Liking for potato chips (p < 0.01), grapes (p < 0.05), cherry tomatoes (p < 0.001), and fruit punch (p < 0.001) was positively associated with amount consumed, but no associations were found between liking and intake of other meal items. For the majority of meal items, test-retest reliability of liking and intake were significant (ranging from 0.34 for cookies to 0.93 for tomatoes). At a multi-component meal, children's hedonic ratings were both reliable and modestly predictive of subsequent intake, and the relationships were stronger for lower energy, less well-liked foods. Rather than eating what they like, these data are more consistent with the notion that children do not eat what they dislike.

Automated summary

Liking plays a primary role in determining children's food intake. This study found a positive relationship between liking and intake for potato chips, grapes, cherry tomatoes, and fruit punch at a meal. The test-retest reliability of liking and intake was significant for most meal items, with stronger relationships observed for less well-liked foods with lower energy content.