Many dietary interventions focus on self-administered reduction in energy intake through the use of calorie-based dietary prescriptions. This approach of calorie counting assumes familiarity of users with calorie content of various foods. There is very little known about the ability of individuals to estimate calorie amounts of food or what factors influence accuracy. This study assessed the accuracy of calorie perceptions in individual food items and mixed meals, and measured body composition. We hypothesized that individuals with higher percent body fat would be less accurate at estimating calories in both visual food displays and written calorie perception questions.
Participants responded to a cross-sectional survey by estimating the calories for a list of individual ingredients and foods, as well as visual food displays. Bioelectrical impedance was used to measure body composition. Spearman correlations were used to analyze associations between calorie perception accuracy and percent body fat.
A total of 451 participants (18-81y) completed the survey (females = 302; Hispanic = 326). We found positive correlations between percent body fat and inaccuracy of response in 22 out of 36 items (R=0.090-0.211; p=0.03-0.001). The inaccuracy of the responses ranged from 0 to 66 times higher than the accurate value. All other items had a positive correlation but were not significant. The strongest correlations between percent body fat and inaccuracy were for a gram carbohydrate, protein, and alcohol (r=0.21-0.19, p<0.001), and for white wine (r=0.19, p<0.001). There were no significant correlations for items such as mac & cheese, enchilada plate, and a chicken drumstick.
We identified an increase in the inaccuracy of calorie perceptions as percent body fat increased in our sample. Further investigation is needed to determine if the accuracy of calorie perceptions is associated with success in weight management efforts.