Tracking nutritional and quality changes in frozen pork: A 12-month study using 7 categories of meat parameters and VIS/NIR spectroscopy

Abstract

Frozen pork stocks are critical for stabilizing food security and prices, but assessing nutritional and physicochemical changes during freezing remains challenging. This study conducted a 12-month frozen storage experiment at −20 °C on 50 pigs' longissimus muscles, quantifying changes in 62 meat quality components in 7 categories, including thawing loss, myoglobin, fatty acids, amino acids, and deterioration indicators. The Prophet model suggests pork retains good quality for 1–2 years under proper freezing, with significant deterioration appearing after 4 years. Phenotypic correlations across dimensions reveal key interactions and abnormalities, such as arachidonic acid (C20:4n6) and methionine (Met) deviating from their respective clusters. Using VIS/NIRS and nine machine learning algorithms, a storage time classification model achieved 85.8 % accuracy with SpecimIQ transmission spectra and AdaBoost. These findings demonstrate the potential of VIS/NIRS in food safety and processing, providing a practical solution for precise storage duration identification and valuable insights for managing frozen pork reserves.