Abstract: This study aimed to scientifically evaluate the main nutritional ingredients of infant formula in order to provide a technical basis for the development and optimization of new infant foods. Gas chromatography, high performance liquid chromatography and ion chromatography were used to detect the three macronutrients proteins, fats and carbohydrates as well as fatty acids, amino acids, lactose and oligosaccharides in infant formula. From the multi-index system, six principal components were extracted and the Mahalanobis distance (MD) was used to measure the nutritional coincidence degree between infant formula and breast milk, and a comprehensive nutritional evaluation was carried out. The results showed that: 1) The compositions and contents of the macronutrients in nine infant formula samples were similar to those in breast milk, but there was a significant difference in some individual components (P < 0.05); 2) Among the tested infant formula samples, goat milk sample D scored highest for Mahalanobis distance (MD = 90.15), followed by goat milk sample F (MD = 86.96), and bovine milk samples H and I showed lower MD scores (82.32 and 82.11, respectively). Accordingly, MD scores for goat milk infant formula were generally higher than those for bovine milk infant formula; 3) MD scores were most affected by fatty acids, and the effect of carbohydrates was the least. The interaction between various nutrients also had an effect. This method can be considered a scientific and reasonable method for the nutritional evaluation of infant formula using multiple nutritional indicators.

Key words: infant formula, principal component analysis, Mahalanobis distance, macronutrients