Reporting Accuracy of Energy Intake Among Exclusively Formula-Fed Infants (P11-097-19)

Assessment of diet is important for understanding nutrition-related health outcomes. Like all methods, diet assessment is subject to error. Error, and in turn reporting accuracy, have been studied extensively in the adult literature but to a lesser extent in the pediatric literature. In the present study, we utilized data on energy intake (EI) and total energy expenditure (TEE) in a racially diverse and contemporary cohort of healthy exclusively formula-fed newborns to determine EI reporting accuracy. In 0.75-month-old term infants (n = 113; 62% black), we measured TEE using the doubly labeled water method and EI using the 3-day weighed bottle method with diet records. Mothers were provided with pre-weighed bottles from which to feed their infants and bottles were weighed by study staff upon return; diet records were analyzed using Nutrient Data Software for Research. Since energy is needed for growth, the estimated energy requirement (EER) was calculated using TEE + energy cost of tissue deposition (175 kcal/d). EI was defined as the mean energy intake over the 3-day period and reporting accuracy was defined as the ratio of EI: EER, which should be equivalent to 1.0 in healthy infants. The cutoff range for implausible reporters was: greater than or less than 1.0 –2*the square root of the average within subject coefficient of variation for EI (12.5%)/100 and TEE (8.2%)/100. EI: EER ratios within the cutoff range (i.e., 0.778-1.222) were categorized as plausible reports, and those below or above the cutoff categorized as non-plausible under-reports and over-reports of EI. The mean EI: EER was 1.037, with a minimum and maximum of 0.360 and 1.907, respectively. Approximately half (49%) of the ratios were deemed plausible, whereas the remainder were implausible under reports (24%) or over reports (27%). Among plausible reports, the mean EI: EER ratio was 0.990 with a minimum and maximum of 0.778 and 1.217, respectively. Despite using the weighed bottle method, only half of the EI reports were plausible. Future research needs to optimize the weighed bottle method to improve accuracy among newly parturient mothers, and consider the accuracy of reported EI when evaluating the impact of early diet on health outcomes. NIH Grants HD072307 and HD094908.