Ingle Prashant D, Christian Roney, Purohit Piyush, Zarraga Veronica, Handley Erica, Freel Keith, Abdo Saleem
Herbalife International of America Inc., Quality Control, 990 West 190th St, Torrance, CA 90501, USA.
J AOAC Int. 2016 Mar-Apr;99(2):360-3. doi: 10.5740/jaoacint.15-0115. Epub 2016 Mar 8.
Protein is a principal component in commonly used dietary supplements and health food products. The analysis of these products, within the consumer package form, is of critical importance for the purpose of ensuring quality and supporting label claims. A rapid test method was developed using near-infrared (NIR) spectroscopy as a compliment to current protein determination by the Dumas combustion method. The NIR method was found to be a rapid, low-cost, and green (no use of chemicals and reagents) complimentary technique. The protein powder samples analyzed in this study were in the range of 22-90% protein. The samples were prepared as mixtures of soy protein, whey protein, and silicon dioxide ingredients, which are common in commercially sold protein powder drink-mix products in the market. A NIR regression model was developed with 17 samples within the constituent range and was validated with 20 independent samples of known protein levels (85-88%). The results show that the NIR method is capable of predicting the protein content with a bias of ±2% and a maximum bias of 3% between NIR and the external Dumas method.
蛋白质是常用膳食补充剂和保健食品中的主要成分。对这些产品在消费者包装形式下进行分析，对于确保质量和支持标签声明至关重要。开发了一种快速检测方法，使用近红外（NIR）光谱作为对当前通过杜马斯燃烧法测定蛋白质的补充。发现近红外方法是一种快速、低成本且绿色（不使用化学物质和试剂）的补充技术。本研究中分析的蛋白粉样品的蛋白质含量范围为22%至90%。样品制备为大豆蛋白、乳清蛋白和二氧化硅成分的混合物，这些成分在市场上商业销售的蛋白粉混合饮料产品中很常见。利用17个在成分范围内的样品建立了近红外回归模型，并用20个已知蛋白质水平（85%至88%）的独立样品进行了验证。结果表明，近红外方法能够预测蛋白质含量，近红外法与外部杜马斯法之间的偏差为±2%，最大偏差为3%。
