Children's body mass index percentile (BMI%), social desirability score (SDS) and dietary recall accuracy for energy

Abstract

We examined whether 4th‐grade children's dietary recall accuracy for energy varied by BMI% and SDS. Each of 374 children (96% Black) was observed eating 2 school meals (breakfast, lunch) and interviewed to obtain a 24‐hour recall that included the 2 observed school meals. At the end of the interview (after the recall), children responded to 14 social desirability questions. We weighed and measured children, and calculated BMI%. We quantified accuracy of recalling energy intake at school meals by calculating conventional report rates (total reported energy/total observed energy) and the 2 components of report rate ‐ correspondence rate (total correctly reported energy/total observed energy) and inflation ratio (total intruded energy/total observed energy). For each measure, ANOVA was conducted with sex, BMI% x sex, SDS x sex, and BMI% x SDS x sex in the model; results were adjusted for other covariables. Report rates increased (indicating less underreporting) as BMI% and SDS decreased (2 p values = 0.001). Correspondence rates increased (indicating better accuracy) as SDS decreased (p = 0.013). Inflation ratios decreased (indicating better accuracy) as BMI% increased (p < 0.001). Report rates alone do not provide a complete picture of the relationship of children's dietary recall accuracy for energy with BMI% and SDS; BMI% and SDS appear to be related to separate aspects of recall accuracy. Funded by R01 HL074358.