Our aim was to assess the use of the Remote Food Photography Method to accurately estimate the weight of infant powdered formula before reconstitution among the standard serving sizes.
For each serving size (1 scoop, 2 scoops, 3 scoops, and 4 scoops), a set of seven test bottles and photographs were prepared as follow: recommended gram weight of powdered formula of the respective serving size by the manufacturer; three bottles and photographs containing 15%, 10%, and 5% less powdered formula than recommended; and three bottles and photographs containing 5%, 10%, and 15% more powdered formula than recommended (n=28). Ratio estimates of the test photographs as compared to standard photographs were obtained using standard Remote Food Photography Method analysis procedures. The ratio estimates and the US Department of Agriculture data tables were used to generate food and nutrient information to provide the Remote Food Photography Method estimates.
Equivalence testing using the two one-sided t tests approach was used to determine equivalence between the actual gram weights and the Remote Food Photography Method estimated weights for all samples, within each serving size, and within underprepared and overprepared bottles.
For all bottles, the gram weights estimated by the Remote Food Photography Method were within 5% equivalence bounds with a slight underestimation of 0.05 g (90% CI −0.49 to 0.40; P<0.001) and mean percent error ranging between 0.32% and 1.58% among the four serving sizes.
The maximum observed mean error was an overestimation of 1.58% of powdered formula by the Remote Food Photography Method under controlled laboratory conditions, indicating that the Remote Food Photography Method accurately estimated infant powdered formula.