2.9. Bias

The food intake records from the average of three 24 h recalls were reviewed to determine under- and over-reporting using Goldberg’s method [29,30]. The under- and over-reporting of dietary intake were identified by comparing the ratio of reported energy intake (EIrep) to the predicted basal metabolic rate (BMRest) with the 95% upper and lower confidence limit of physical activity level (PAL):

The mean of energy intake from the three 24-h recalls for each student was used as estimated energy intake. The predicted basal metabolic rate was calculated for each child using established equations adjusted for weight, height, relevant sex, and age groups in consideration the context of overweight and obesity in the study population [31,32]. The 95% lower and upper limit were calculated with the physical activity level was 1.55 due to the evidence of the low level of physical activity among fifth-grade children in Ho Chi Minh City [15] and the average within-subject variation in intake (S) calculated by the equation below, with suggested values for within-subject variation in energy intake (CV_wEI), within-subject variation in repeated BMR measurement (CV_wB), total between-subject variation in PAL (CV_tP), and the number of days of dietary assessments were 23%, 8.5% and 15%, 3 days, respectively [33].