2025 Volume 74 Issue 4 Pages 361-375
To accurately quantify the phospholipids in infant formula by subclass, we developed an analytical method using phosphorus-31 nuclear magnetic resonance spectroscopy (31P NMR). We performed heated extraction method using a mixture of ethanol and water to extract phospholipids from infant formula and replace the highly toxic chloroform traditionally used for extraction. In the 31P NMR measurement, we also avoided using chloroform by dispersing the extracts in surfactants with a strong affinity for phospholipids. Although polar lipids in milk are characterized by a high content of sphingomyelin, the separation of sphingomyelin and phosphatidylethanolamine signals was insufficient to accurately determine their signal areas. To overcome this issue, we applied and evaluated two different methods, integration and deconvolution, for calculating the signal areas. During method validation in a spiked recovery test, the deconvolution method gave a recovery rate closer to 100% than the integration method. The main phospholipid subclasses found in infant formula were phosphatidylcholine, phosphatidylinositol, phosphatidylserine, phosphatidylethanolamine, and sphingomyelin. However, when the formula contained soy lecithin, additional soy-derived phospholipids, such as phosphatidic acid, were detected. Using equipment with a phosphorus resonance frequency of 202 MHz and a measurement time of approximately 4 h, the quantification limit was 5 mg/100 g. The developed method will be useful for analysis of phospholipids in infant formula.
